/3-- :p-

1
.>
.*.-- en-

-+ & [-71; &v

f f - -. . - I&- -.
'
'I ;
('A-
. I

; < ,2
I
J
/
1,;
'

Ik

BIOMEDICAL
DEVICE
TECHNOLOGY
\
BIOMEDICAL DEVICE TECHNOLOGY
 ABOUT THE AUTHOR

Anthony Y.K. Chan graduated in Electrical Engineering (B.Sc.

Hon.) from the University of Hong Kong in 1979, completed a
M.Sc. degree in Engineering from the same university, and worked
for a number of years as a project engineer in electrical instrumen-
tations, control, and systems. In 1987, he completed a Master of
Engineering Degree (M.Eng.) in Clinical Engineering from the
University of British Columbia in Canada. He also holds a
Certificate in Hospital Services Management from the Canadian
Healthcare Association. Anthony was the director and manager of
biomedical engineering in a number of Canadian acute care hospi-
tals. He is currently the Program Head of the Biomedical Engineer-
ing Technology Program at the British Columbia Institute of
Technology and Adjunct Professor of the Biological and Chemical
Engineering Department of the University of British Columbia. He
is a Professional Engineer, a Chartered Engineer, and a Certified
Clinical Engineer.
BIOMEDICAL DEVICE
TECHNOLOGY
Principles and Design

BY

ANTHONY Y. K. CHAN, MSc, MEng, PEng, CCE

CHARLES C THOMAS P U B L I S H E R , LTD.

S'ringf?eld Illinois U.S.A.
 Published and Distributed 27troughout the World by

CHARLES C THOMAS PUBLISHER, LTD.

2600 South First Street
Springfield, Illinois 62704

This book is protected by copyright. No part of

it may be reproduced in any manner without written
permission from the publisher. All rights reserved.

O 2008 by CHARLES C THOMAS PUBLISHER, LTD.

ISBN 978-0-398-07699-3(hard)
ISBN 978-0-398-07700-6 (paper)

Library of Congress Catalog Card Number: 2006050066

With THOMAS BOOKS carefil attention is given to all detaibs of manufacturing

and design. It is the Pub1isher"sesire to present books that are satisfactory as to their
physical qualities and artistic possibilities and appropriate for their particular use.
THOMAS BOOKS will be true to those laws ofqualily that mure a good name
and good will.

Printed in the United States of America

SM-R-3

Library of Congress Cataloging-in-PublicationData

Chan, Anthony Y. K.
Biomedical device technology : principles and design / by Anthony Y. K.
Chan.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-398-07699-3 -- ISBN 978-0-398-07700-6 (pbk.)
1. Medical instruments and apparatus. 2. Medical technology. 3.
Biomedical engineering. I. Title.
 To my w i f , Elaine
and
my daughters, Victoria and Tzflany
 PREFACE

F or many years, the tools available to physicians were limited to a few sim-
ple handpieces such as stethoscopes, thermometers, and syringes; med-
ical professionals primarily relied on their senses and skills to perform diag-
nosis and disease mitigation. Today, diagnosis of medical problems is heavi-
ly dependent on the analysis of information made available by sophisticated
medical machineries such as electrocardiographs, ultrasound scanners, and
laboratory analyzers. Patient treatments often involve specialized equipment
such as cardiac pacemakers and electrosurgical units. Such biomedical in-
strumentations play a critical and indispensable role in modern medicine.
In order to design, build, maintain, and effectively deploy medical de-
vices, one must understand not only their design and construction but also
how they interact with the human body. This book provides a comprehen-
sive approach studying the principles and design of biomedical devices as
well as their applications in medicine. It is written for engineers and tech-
nologists who are interested in understanding the principles, design, and
applications of medical device technology. The book is also intended to be
used as a textbook or reference for biomedical device technology courses in
universities and colleges.
The most common reason of medical device obsolescence is changes in
technology. For example, vacuum tubes in the 1960s, discrete semiconduc-
tors in the 1970s, integrated circuits in the 1980s, microprocessors in the
1990s, and networked multiprocessor software-driven systems in today's
devices. The average life span of medical devices has been diminishing; cur-
rent medical devices have a life span of about 5 to 7 years. It is unrealistic to
write a book on medical devices and expect that the technology described
will remain current and valid for years. On the other hand, the principles of
medical device applications, the origins of physiological signals and their
methods of acquisition, and the concepts of signal analysis and processing
will remain largely unchanged. This book focuses on the functions and prin-
ciples of medical devices (which are the invariant components) and uses spe-
cific designs and constructions to illustrate the concepts where appropriate.

vii
viii Biomedical Device Technology: Princ$kes and Design

The first part of this book discusses the fundamental building blocks of
biomedical instrumentations. Starting from an introduction of the origins of
biological signals, the essential functional building blocks of a typical med-
ical device are studied. These functional blocks include electrodes and trans-
ducers, biopotential amplifiers, signal conditioners and processors, electrical
safety and isolation, output devices, and visual display systems. The next sec-
tion of the book covers a number of biomedical devices. Their clinical appli-
cations, principles of operations, functional building blocks, special features,
performance specifications, as well as common problems and safety precau-
tions are discussed. Architectural and schematic diagrams are used where
appropriate to illustrate how specific device functions are being implement-
ed.
Due to the vast variety of biomedical devices available in health care, it
is impractical to include all of them in a single book. This book selectively
covers diagnostic and therapeutic devices that are either commonly used or
whose principles and design represent typical applications of the technology.
To limit the scope, medical imaging equipment and laboratory instrumenta-
tions are excluded from this book.
Three Appendices are included at the end of the book. These are
appended for those who are not familiar with these concepts yet an under-
standing in these areas will enhance the comprehension of the subject mat-
ters in the book. They are: A-1. A Primer on Fourier Analysis; A-2.
Overview of Medical Telemetry Development; and A-3. Medical Gas
Supply Systems.
I would like to take the opportunity to acknowledge Euclid Seeram, who
encouraged and inspired me to embark in writing, and Michael Thomas for
agreeing to publish and giving me the extra time to finish this book.

Anthony Y, K Chan
 CONTENTS

Page
Preface . . . . . . . . . . . . . . . . . . . . . . . . .VU
Chapter

1. OVERVIEW OF BIOMEDICAL INSTRUMENTATION . . . . . . .3

2. CONCEITS IN SIGNAL MEASUREMENT.
PROCESSING AND ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . .28
BIOMEDICAL TRANSDUCERS
3 . FUNDAMENTALS OF BIOMEDICAL TRANSDUCERS . . . . . .42
4 . PRESSURE AND FORCE TRANSDUCERS . . . . . . . . . . . . . . . . .53
5. TEMPERATURE TRANSDUCERS . . . . . . . . . . . . . . . . . . . . . . . 65
.
6. MOTION TRANSDUCERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87
7. FLOW TRANSDUCERS . . . . . . . . . . . . . . . . . . . . . . . . . . .-95
8. OPTICAL TRANSDUCERS . . . . . . . . . . . . . . . . . . . . . . . . . 108
9. ELECTROCHEMICAL TRANSDUCERS . . . . . . .125
10. BIOPOTENTIAL ELECTRODES . . . . . . . . . . . . . . . . . . . . .148
FUNDAMENTAL BUILDING BLOCKS
OF MEDICAL INSTRUMENTATION
11. BIOPOTENTIAL AMPLIFIERS . . . . . . . . . . . . . . . . . . . . . . . . . . .158
12. ELECTRICAL SAFETY AND SIGNAL ISOLATION . . . . . . . . .182
 Biomedical Deuice Technology: Princ$les and Design

13. MEDICAL WAVEFORM DISPLAY SYSTEMS ..............202

MEDICAL DEVICES
14. PHYSIOLOGICAL MONITORING SYSTEMS . . . . . . . . . . . . .223
15. ELECTROCARDIOGRAPHS ............................. 239
16. ELECTROENCEPHALOGRAPHS . . . . . . . . . . . . . . . . . . . . . . . .262
17. ELECTROMYOGRAPHY AND EVOKED POTENTIAL
STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .281
18. INVASIVE BLOOD PRESSURE MONITORS ..............294
19. NONINVASIVE BLOOD PRESSURE MONITORS . . . . . . . . . .312
20. CARDIAC OUTPUT MONITORS ........................323
21. CARDIAC PACEMAKERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
.
22 . CARDIAC DEFIBRILLATORS ...........................355
23. INFUSION DEVICES ...................................371
24. ELECTROSURGICAL UNITS ............................391
25. RESPIRATION MONITORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
26 . MECHANICAL VENTILATORS ..........................422
27. ULTRASOUND BLOOD FLOW DETECTORS .............435
28 . FETAL MONITORS .....................................443
29. INFANT INCUBATORS, WARMERS, AND
PHOTOTHERAPY LIGHTS ............................449
30. BODY TEMPERATURE MONITORS . . . . . . . . . . . . . . . . . . . . .457
31. PULSE OXIMETERS ....................................469
32. END-TIDAL CARBON DIOXIDE MONITORS .............480
33. ANESTHESIA MACHINES ..............................486
34 . DIALYSIS EQUIPMENT ................................499
35 . MEDICAL LASERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .520
36 . ENDOSCOPIC VIDEO SYSTEMS ........................536
 Contents

APPENDICES
A.1 . A PRIMER ON FOURIER ANALYSIS .....................547
A.2 . OVERVIEW OF MEDICAL TELEMETRY
DEVELOPMENT .....................................551
A.3 . MEDICAL GAS SUPPLY SYSTEMS ....................... 555
Index .......................................................5 59
 Chapter 1

OVERVIEW OF BIOMEDICAL
INSTRUMENTATION

Define "medical device."

Analyze biomedical instrumentation using a systems approach.
Explain the origin and characteristics of biopotentials and common
physiological signals.
Explain the importance and approaches of human factor engineering in
medical device design.
List common input, output, and control signals of medical devices.
Identify the special constraints encountered in the design of biomedical
devices.
Define biocompatibility and list common biomaterials used in medical
devices.
Explain the tissue responses and approaches to achieve biocompatibil-
ity.
Identify the basic functional building blocks of medical instrumenta-
tion.

CHAPTER CONTENTS

1. Introduction
2. Classification of Medical Devices
3. Systems Approach
4. Origins of Biopotentials
5. Physiological Signals
 Biomedical Device Technology: Princ$les and Design

6. Human Machine Interface

7. Input, Output, and Control Signals
8. Constraints in Biomedical Signal Measurements
9. Concepts on Biocompatibility
10. Functional Building Blocks of Medical Instrumentation

INTRODUCTION

Medical devices come with different designs and complexity. They can
b e as simple as a tongue depressor, as compact as a rate-responsive demand
pacemaker, or as sophisticated as a surgical robot. Although most medical
devices use similar technology as their commercial counterparts, there are
many fundamental differences between devices used in medicine and
devices used in other applications. This chapter will look at the definition of
medical devices and the characteristics that differentiate a medical device
from other household or commercial products.
According to the United States Food and Drug Administration (FDA), a
"medical device" is defined as:
"an instrument, apparatus, implement, machine, contrivance, implant, in
vitro reagent, or other similar or related article, including a component part,
or accessory which is:
recognized in the official National Formulary, or the United States
Pharmacopoeia, or any supplement to them,
intended for use in the diagnosis of disease or other conditions, or in the
cure, mitigation, treatment, or prevention of disease, in man or other ani-
mals, or
intended to affect the structure or any function of the body of man or
other animals, and which does not achieve any of its primary intended
purposes through chemical action within or on the body of man or other
animals and which is not dependent upon being metabolized for the
achievement of any of its primary intended purposes."
A "medical device" is similarly defined in the Canadian Food and Drugs
Act, as:
"Any article, instrument, apparatus or contrivance, including any compo-
nent, part or accessory thereof, manufactured, sold or represented for use
in:
(a) the diagnosis, treatment, mitigation or prevention of a disease, disorder
or abnormal physical state, or the symptoms thereof, in humans or ani-
mals;
(b) restoring, correcting or modifying a body function, or the body structure
of humans or animals;
 Overview of Biomedical Instrumentation

(c) the diagnosis of pregnancy in humans or animals; or

(d) the care of humans or animals during pregnancy, and at, and after, birth
of the offspring, including care of the offspring, and includes a contra-
ceptive device but does not include a drug."
Apart from the obvious, it is clear from these definitions that in vitro
diagnostic products such as medical laboratory instruments are medical
devices. Furthermore, accessories, reagents, or spare parts associated with a
medical device are also considered to be medical devices. An obvious exam-
ple of this is the electrodes of a heart monitor. Another example, which may
not be as obvious, is the power adapter to a medical device such as a laryn-
goscope. Both of these accessories are considered as medical devices and are
therefore regulated by the premarketing and postmarketing regulatory con-
trols.

CLASSIFICATION OF MEDICAL DEVICES

There are many different approaches to classify or group medical

devices. Devices can be grouped by their functions, their technologies, or
their applications. A description of some common classification methods fol-
lows.

Classified by Functions
Grouping medical devices by their functions is by far the most common
way to classify medical devices. Devices can be separated into two main cat-
egories: diagnostic and therapeutic.
Diagnostic devices are used to determine physical signs and diseases
and/or injury without alteration of the structure and function of the biologi-
cal system. However, some diagnostic devices may alter the biological sys-
tem to a certain extent due to their applications. For example, a real-time
blood gas analyzer may require invasive catheters (which puncture the skin
into a blood vessel) to take PC02 measurement. A computer tomography
scanner will impose ionization radiation (transfer energy) on the human
body in order to obtain the medical images.
Diagnostic devices whose functions are to determine the changes of cer-
tain physiological parameters over a period of time are often referred to as
monitoring devices. As the main purpose of this class of devices is trending,
absolute accuracy may not be as important as their repeatability. Examples
of monitoring devices are heart rate monitors used to detect variation of
heart rates during a course of drug therapy, and noninvasive blood pressure
 Biomedical Device Technology:Principles and Design

monitors to watch arterial blood pressure immediately after surgery.

Therapeutic devices are designed to effect structural or functional
changes that lead to improved overall function of the patient. Examples of
such devices are electrosurgical units in surgery, linear accelerators in cancer
treatment, and infusion devices in fluid management therapy. Assistive
devices are a group of devices used to restore an existing function of the
human body. They may be considered a subset of therapeutic devices.
Examples are demand pacemakers to restore normal heart rhythm, hearing
aids to assist hearing, and wheelchairs to enhance mobility of the disabled.
Based on the methods of application, these device classes can be further
subdivided into invasive or noninvasive, automatic or manual.

Classified by Physical Parameters

Medical devices can also be grouped by the physical parameters that
they are measuring. For example, a blood pressure monitor is a pressure
monitoring device, an airway spirometer is a flow measurement device, and
a tympanic thermometer is a temperature-sensing device.

Classified by Principles of Transduction

Some medical devices are grouped according to the types of transducers
used at the patient-machine interface such as resistive, inductive, or ultra-
sonic.

Classified by Physiological Systems

Medical devices may also be grouped by their related human physiolog-
ical systems. Examples of such grouping are blood pressure monitors and
electrocardiographs as cardiovascular devices, and respirators and mechani-
cal ventilators as pulmonary devices.

Classified by Clinical Medical Specialties

In another model, devices are grouped according to the medical special-
ties (such as pediatrics, obstetrics, etc.) in which they are being used. For
example, a fetal monitor is considered as an obstetric device, an x-ray
machine as a radiological device.
 Oueruiew of Biomedical instrumentation

Classified by Risk Classes

For biomedical engineers and regulatory personnel, medical devices are
often referred to by their risk class. Risk classes are created to differentiate
devices by rating their risk level on patients. A device risk classification
determines the degree of scrutiny and regulatory control imposed on the
manufacturers and users by regulatory bodies to ensure their safety and effi-
cacy in clinical use. Table 1-1 shows examples of medical dev.ices in each
risk class under the Canadian Medical Device Regulations. Similar risk clas-
sifications are used in the United States and Europe.

Table 1-1.
- -
Risk Classif~cation
-

Four risk classes-from Class I (lowest risk) to Class IV (highest risk)

Class I conductive electrode gel, band-aids
Class I1 latex gloves, contact lenses
Class 111 IV bags, indwelling catheters
-
Class IV
- .
heart valve implants, defibrillators

SYSTEMS APPROACH

In simple terms, a system is defined as a group of things or parts or

processes working together under certain relationships. A system transforms
a set of input entities to a set of output entities. Within a system there are
aspects, variables, or parameters that mutually act on each other. A closed
system is self-contained on a specific level and is separated from and not
influenced by the environment, whereas an open system is influenced by the
environmental conditions by which it is surrounded. Figure 1-1 shows an
example of a system. The elements within a system and their relationships as
well as the environment can affect the performance of the system. A more
complicated system may contain multiple numbers of subsystems or simple
systems.
In analyzing a large complex system, one can divide the system into sev-
eral smaller subsystems with the output from one subsystem connected to the
input of another. The simplest subsystem consists of an input, an output, and
a process as shown in Figure 1-2. The process that takes the output and feeds
it back to the input in order to modify the output is called a feedback process.
A system with feedback is called a closed-loop system, whereas a system
without any feedback is called an open-loop system. Most systems that we
encounter contain feedback paths and hence are closed-loop systems.
 Biomedical Device Technology: Principles and Design

Environment

Figure 1-1. Typical System.

Input Process b Output

A +

Feedback

Figure 1-2. Basic Subsystem.

Listening to radio is an example of a simple closed-loop system. The

input to the system is the radio broadcast in the form of an electromagnetic
wave that is received by the radio. The radio processes the received signal
and produces the audible sound such as music. If the music (output) is not
loud enough, you turn up the volume to increase the sound level. In doing
this, you become the feedback process that analyzes the loudness of the
music and produces the action to turn up the volume.
The systems approach is basically a generalized technique to understand
organized complexity. It provides a unified framework or a way of thinking
about the systems and can be developed to handle specific problems. In
order to solve a problem, one must look at all components within the system
and analyze the input and output of each subsystem in view to isolate the
problem and establish the relationships of the problem with respect to each
component in the system.
Using block diagrams to analyze complex devices is an application of the
systems approach. Figure 1-3 shows a compact disc (CD) music player sys-
tem. The input to the CD player is the musical CD, the output is sound (or
 Overview of Biomedical Instrumentation

music), and the feedback is the listener who will replace the CD when it has
finished playing or turn down the volume if it is too loud. If the CD player
is not working properly, one may buy a new one and discard the malfunc-
tioning unit.

A b CD Player b Music

Listener

Figure 1-3. CD System.

The CD player can be divided into its functional blocks as shown in

Figure 1-4. One may be able to troubleshoot and isolate the problem to one
of the functional blocks. In this case, it will be cheaper just to replace the mal-
functioning block. For example, if the speakers are not working, it may be
more economical to get a pair of replacement speakers than to replace the
entire CD player.

CD 7andReader
Decoder Amplifier --b Speakers .--b Music

Listener 4

Figure 1-4. CD Player Functional Block.

Similarly, a complex biomedical device can be broken down into its

functional building blocks. Figure 1-5 shows a block diagram of an electro-
cardiogram (ECG) system. The input to the device is the biopotential from
the heart activities. The electrodes pick up this tiny electrical signal from the
patient and send it to the amplifier block to increase the signal amplitude.
The amplified ECG signal is then sent to the signal analysis block to extract
information such as the heart rate. Finally, the ECG signal is sent to the out-
 Biomedical Device Technology: Princ$les and Design

put block such as a paper chart recorded to produce a hard copy of the ECG
tracing. These blocks can be further subdivided, eventually down to the indi-
vidual component level. Note that the cardiology technologist is also consid-
ered a part of the system. He or she serves as the feedback loop by monitor-
ing the output and modifying the input.
When analyzing or troubleshooting a medical device, it is important to
understand the functions of each building block, and what to expect from the
output when a known input is applied to the block. Furthermore, medical
devices are, in most cases, conceptualized, designed, and built from a com-
bination of functional building blocks or modules.

Patient -+ Electrode -, Amplifier -,.

Signal
Analyzer
-b Recorder --+ ECG
A Printout

ECG Technologist 4

Figure 1-5. ECG Block Diagram.

ORIGINS OF BIOPOTENTIALS

The source of electrical events in biological tissue is the ions in the elec-
trolyte solution, as opposed to the electrons in electrical circuits. Biopotential
is an electrical voltage caused by a flow of ions through biological tissues. It
was first studied by Luigi Galvani, an Italian physiologist and physicist, in
1786. In living cells, there is an ongoing flow of ions (predominantly sodium-
Na+, potassium-K+ and chloride-C1) across the cell membrane. The cell
membrane allows some ions to go through readily but resists others. Hence
it is called a semipermeable membrane.
There are two fundamental causes of ion flow in the body: diffusion and
drift. Fick's law states that if there is a high concentration of particles in one
region and they are free to move, they will flow in a direction that equalizes
the concentration; the force that results in the movement of charges is called
diffusion force. The movement of charged particles (such as ions) that is due
to the force of an electric field (static forces of attraction and repulsion) con-
stitutes particle drift. Each cell in the body has a potential voltage across the
cell membrane known as the single-cell membrane potential.
Under equilibrium, the net flow of charges across the cell membrane is
 Overview of Biomedical Instrumentation

zero. However, due to an imbalance of positive and negative ions internal

and external to the cell, the potential inside a living cell is about -50 mV to
-100 mV with respect to the potential outside it (Figure 1-6). This membrane
potential is the result of the diffusion and drifi of ions across the high resis-
tance but semipermeable cell membrane, predominantly sodium [Na'] and
potassium [K'] ions moving in and out of the cell. Because of the semiper-
meable nature of the membrane, Na+ is partially restricted from passing into
the cell. In addition, a process called the sodium-potassium pump moves
sodium ions at two to five times the rate out of the cell than potassium ions
into the cell. However, in the presence of diffusion and drift, an equilibrium
point is established when the net flow of ions across the cell's membrane
becomes zero. As there are more positive ions (Na') moved outside the cells
than positive ions (K') moved into the cell, under equilibrium, the inside of
the cell is more negative than the outside. Therefore, the inside of the cell is
negative with respect to the outside. This is called the cell's resting potential,
, which is typically about -70 mV.

\ -70 mV
4 \
I \
I 1

\.
I 1
I I
Living Cell ,
0
I Reference

**-/
Figure 1-6. Cell Membrane Potential.

If the potential across the cell membrane is raised, for example by an

a1 stimulation, to a level that exceeds the threshold, the ,permeability
e cell membrane will change, causing a flow of Na' ions into the cell.
o inrush of positive ions will create a positive change in the cell's mem-
potential to about 20 mV to 40 mV more positive than the potential
e the cell. This action potential lasts for about 1 to 2 milliseconds. As
the action potential exists, the cell is said to be depolarized. The
ane potential will drop eventually as the sodium-potassium pump
es the cell to its resting state (-70 mV). This process is called repo-
on and the time period is called the refractory period. During the
tory period, the cell is not responsive to any stimulation.
Ther
v events of depolarization and repolarization are shown in Figure 1-7.
 Biomedical Device Technology: Principles and Design

The rise in the membrane potential from its resting stage (when stimulated)
and return to the resting state is called the action potential. Cell potentials
form the basis of all electrical activities in the body, including such activities
as the electrocardiogram (ECG), electroencephalogram (EEG), electrooculo-
gram (EOG), electroretinogram (ERG), and electromyogram (EMG).

Potential
A

+20 mV
b
Time

-70 mV

2 msec
Figure 1-7. Action Potential.

When a cell is depolarized (during which the membrane potential

changes from negative to positive), the cells next to it may be triggered into
depolarization. This disturbance is propagated either to adjacent cells, result-
ing in the entire tissue becoming depolarized (in an entire motor group), or
along the length of the cell from one cell to the next (in a single motor unit
or a nerve fiber).
In most biopotential signal measurements, unless one is using a needle
electrode to measure the action potential of a single cell, the measured signal
is the result of multiple action potentials from a g r ~ u pof cells or tissue. The
amplitude and shape of the biopotential are largely dependent on the loca-
tion of the measurement site and the signal sources. Furthermore, the biopo-
tential signal will be altered as it propagates along the body tissue to the sen-
sors. A typical example of biopotential measurement is measuring electrical
heart activities using skin electrodes (electrocardiogram or ECG). Figure 1-8
shows a typical ECG waveform showing the electrical heart potential when
a pair of electrodes is placed on the chest of the patient. This biopotential,
which is the result of all action potentials from the heart tissue transmitted to
the skin surface, is very different in amplitude and shape from the action
potential from a single cell shown in Figure 1-7. In addition, placing the skin
 Ovenliew of Biomedical Instrumentation

electrodes at different locations on the patient will produce very different

looking ECG waveforms.

Figure 1-8. Typical ECC Obtained f r m Skin Electrodes.

PHYSIOLOGICAL SIGNALS

Biopotentials represent a substantial proportion of human physiological

signals. In addition, there are other forms of physiological signals, such as
pressure and temperature, all of which contain information that reflects the
well-being of an individual. Monitoring and analyzing such parameters is of
interest to medical professionals. Different physiological signals have differ-
ent characteristics. Some physiological signals are very small compared with
other background signals and noise; some change rapidly during the course
of their measurement. Therefore, different transducers with matching char-
acteristics are necessary in medical devices to accurately measure these sig-
nals. Table 1-2 shows some examples of common physiological signals; their
characteristics and examples of the transduction techniques used to capture
these signals are also listed. The range and bandwidth quoted in the list are
nominal values, which may not include some extreme cases. An example is
' severe hypothermia, in which the body temperature can become many
degrees below 32OC.
An example of a physiological signal measurement is the electrocardio-
@am. When skin electrodes are placed on the surface of a patient's chest,
they pick up a small electrical potential at the skin surface from the activities
ofthe heart. If one plots this potential against time, this is called an electro-
cardiogram. An example of an electrocardiogram is shown in Figure 1-8.
The spike is called the R-wave, which coincides with the contraction phase
idthe ventricles. The time interval between two adiacent R-waves re~resents
l&e heart cycle. The amplitude and the shape of &e ECG signal depend on
' EL'
 Biomedical Device Technology: Princ9les and Design

the physiological state of the patient as well as the locations and the types of
electrodes used. From Table 1-2, the amplitude of the R-wave may vary
from 0.5 to 4 mV, and the ECG waveform has a frequency range or band-
width from 0.01 to 150Hz.
There are many more physiological signals than those listed in the table.
While some are common parameters in clinical settings (e.g., body tempera-
ture), others are used sparingly (e.g., electroretinogram).

Table 1-2.
- - - -
Characteristics
- - -
of
- -
Common -
Physiological Parameters
- - --- - - - -
Physiological Physical Units Signal Frequenq Measurement Method
Parameters and Range of Range of or Transducer Used
- - -
Measurement - -
Bandwidth
- - - - - -

Blood Flow 1 to 300 mL/s 0 to 20 Hz Ultrasound

Doppler flowmeter
Blood Pressure-Arterial 20 to 400 mmHg 0 to 50 Hz Sphygmomanometer
Blood Pressure-Venous 0 to 50 mmHg 0 to 50 Hz Semiconductor
strain gauge
Blood pH 6.8 to 7.8 Oto2Hz pH electrode
Cardiac Output 3 to 25 Wmin 0 to 20 Hz Thermistor
(thermodilution)
Electrocardiography 0.5 to 4 mV 0.01 to 150 Hz Skin electrodes
(ECG)
Electroencephalography 5 to 300 pV Scalp electrodes
(EEG)-scalp
EEG-brain surface 10 to 5,000 pV Cortical or
or depth depth electrodes
Electromyography (EMG) 0.1 to 5 mV Needle electrodes
Nerve Potentials 0.01 to 3 mV Needle electrodes
Oxygen Saturation- 85 to lOO(Yo Differential light
Arterial (noninvasive) absorption
Respiratory Rate 5 to 25 Skin electrodes
breathlmin (impedance
pneumography)
Tidal Volume Spirometer
Temperature-Body
- - - - - -
to 40°C
32- --
0 to 0.1 Hz
- - -
Thermistor - -

HUMAN-MACHINE INTERFACE

A medical device is designed to assist clinicians to perform certain diag-

nostic or therapeutic functions. In fulfilling these functions, a device inter-
faces with the patients as well as the clinical users. Figure 1-9 shows the inter-
 Overview of Biomedical Instrumentation

faces between a medical device, the patient, and the clinical staff. For a diag-
nostic device, the physiological signal from the patient is picked up and
processed by the device; the processed information such as the heart rhythm
from an ECG monitor or blood pressure waveform from an arterial blood
line is displayed by the device and reviewed by the clinical staff. For a ther-
apeutic device, the clinical staff will, using the device, apply certain actions
on the patient. For example, a surgeon may activate the electrosurgical hand
piece during a procedure to coagulate a blood vessel. In another case, a
nurse may set up an intravenous infusion line to deliver medication to a
patient.

___-------------__

-
___----- -----____
- -.
I
0
. /--
*---
Environment
C

----. 5
\

r
I
____+ + Clinical
\
\
\ Patient
.
I
x + Staff 0
I
v
*.
--- - - - -----__ _-cC
C
d
4

........................ _C_-----

Figure 1-9. Human-Machine Interface.

These interfaces are important and often critical in the design of bio-
medical devices. An effective patient-machine interface is achieved through
carefully choosing a transducer suitable for the application. For example, an
implanted pH sensor must pick up the small changes in the hydrogen ion
concentration in the blood; at the same time it also must withstand the cor-
rosive body environment, maintain its sensitivity, and be nontoxic to the
patient.
Other than safety and efficacy, human factor is another important con-
sideration in designing medical devices. Despite the fact that human error is
a major contributing factor toward clinical incidents involving medical
devices, human factor is often overlooked in medical device design and in
device acquisitions. The goal to achieve in user-interface design is to improve
efficiency, reduce error, and prevent injury. Human factor engineering is a
systematic, interactive design process that is critical to achieve an effective
user-interface. It involves the use of various methods and tools throughout
the design life cycle. Classical human factor engineering involves analysis of
sensory limitations, perceptual and cognitive limitations, and effector limita-
tions of the device users as well as the patients. Sensory limitation analysis
evaluates the responses of the human visual, auditory, tactile, and olfactory
systems. Perceptual and cognitive limitation analysis studies the nervous sys-
 Biomedical Device Technology: Princ9les and Design

tem's response to the sensory information. Perception refers to how people

identify and organize sensory input; cognition refers to higher-level mental
phenomena such as abstract reasoning, formulating strategies, formation of
hypothesis, et cetera. Effector limitation analysis evaluates the outputs or
responses of the operators (e.g., the reaction time, force-exerting capability,
etc.).
There are three subjects to be focused on in human factor design in med-
ical devices: the user, the patient, and the support staff. The three areas of
limitations described above must be considered in each case (Figure 1-10).

Subjects Limitations

Figure 1-10. Medical Device Human Factor Considerations.

User Focus
For diagnostic devices, users rely on the information from the medical
device to perform diagnosis. The display of information should be clear and
unambiguous. It is especially important in clinical settings, where errors are
often intolerable. In a situation in which visual alarms might be overlooked,
loud audible alarms to alert one to critical events should be available. For
therapeutic devices, ergonomic studies should be carried out in the design
stage to ensure that the procedures could be performed in an effective and
efficient manner. Critical devices should be intuitive and easy to set up. For
example, a paramedic should be able to correctly perform a cardiac defib-
rillation without going through complicated initialization procedures since
every second counts when a patient is in cardiac arrest.
 Overview of Biomedical Instrumentation

A systems approach to analyze human interface related to users should

consider the following:
User characteristics
Operating environment
Human mental status
Task priority
Work flow
Human interface outputs may involve hand, finger, foot, head, eye,
voice, et cetera. Each should be studied to identify the most appropriate
choice for the application. A device should be ergonomically designed to
minimize the strain and potential risk to the users, including long-term health
hazards. For example, a heavy X-ray tube can create shoulder problems for
radiology technologists who spend most of their working days maneuvering
X-ray tubes over patients. Studies show that user fatigue is a major contribu-
tor to user errors. User fatigues include motor, visual, cognitive, and memo-
'Y.
Traditionally, human factor engineering is task-oriented. It examines and
optimizes tasks to improve output quality, reduce time spent, and minimize
the rate of error. Proactive human interface designers tend to be user-cen-
tered, who integrate the physical and mental states of the user into the
design, including the level of fatigue and stress, as well as recruit emotional
feedback. Ideally, a good human interface design will produce a device that
is both user-intuitive and efficient. However, in most cases, there is a balance
and trade-off between the two. An intuitive design is easy to use, that is, a
user can learn to operate the device in a short time. However, the operation
of such a device may not be efficient. An example of such a device is a PACS
(picture archiving and communication system) using a standard computer
mouse as the human machine interface between the user and the PACS. The
mouse is intuitive to most users. However, a radiologist may require going
,through a large number of moves and clicks to complete a single task. On
he other hand, a specially designed, multibutton, task-oriented controller
my be difficult to learn initially but will become more efficient once the
,,dologisthas gotten used to it. Figure 1-11 shows the efficiency-time learn-
,hgcurve of a device by a new user. The learning time for an intuitive device
j~shorter than a specially designed device, but the efficiency is much lower
limcethe user becomes proficient with the specially designed device.

Patient Focus
Traditionally, in designing a medical device, much attention is given to
@e safety and efficacy of the system. However, it is also important to look at
 Biomedical Device Technology: Princ$les and Design

Efficiency

Special designed

Time
Figure 1-1 1. Learning Curve.

the design from the patient's perspectives. A good medical device design
should be aesthetically pleasing to the eye and will not interfere with the nor-
mal routines of the patient. Some examples to illustrate the importance of
human factor design related to patients are:
A model of an infrared ear thermometer looks like a pistol with a trig-
ger. The patient may feel threatened when the clinician points it into
his or her ear and pulls the trigger.
A motorized fan in an infant incubator is too noisy. It disturbs the
sleep of the baby and may even inflict hearing damage.
For a person who requires 24-hour mechanical ventilation, a tra-
cheostomy tube that cannot be concealed properly may affect his (or
her) social life.

Support Staff Focus

In designing a medical device, the ergonomics of maintenance tasks such
as cleaning and servicing are often overlooked. Apart from its desired appli-
cation, a medical device will be handled by many parties during its life cycle.
A device that has difficult-to-assess hollow cavities will have problems in
cleaning and sterilization and hence is not suitable for some medical proce-
dures. Some devices are not service-friendly; many poorly designed devices
require extensive dismantling in order to get access to replacement parts
such as lightbulbs and batteries. Other devices may not have taken into con-
sideration the operating environment, which in most cases will result in
expensive and labor-intensive maintenance. An example is a fan-cooled
device used in a dusty environment. In addition, poor design of accessories
may increase the chance of incorrect assembly, which can impose unneces-
sary risk on the patients.
 Overvieu of Biomedical Instrumentation

INPUT, OUTPUT, AND CONTROL SIGNALS

A simple system has a single input and a single output. When we study
a medical device using the systems approach, the first step is to analyze the
input to the device. In most cases, input signal to biomedical devices are
physiological signals. In order to study the characteristics of the output, one
must understand the nature of the processes that the device applies to the
input. In addition to the main input and output signals, most medical devices
have one or more control inputs (Figure 1-12). These control inputs are used
by the operator to select the functions and control the device. Table 1-3 lists
some examples of input, output, and control signals in biomedical devices.

Input Output

Control
Figure 1-12. Medical Device System.

Table 1-3.
--
Examples of Medical Device InpuVOutput
--

Signal Input
Electrical potential in ECG
Pressure signal in blood pressure monitoring
Heat in body temperature measurement
Carbon dioxide partial pressure in end-tidal C 0 2 monitoring
Device Output
Printout in paper chart recorder
Signal waveform in CRT display
Alarm signal in audible tone
Heat energy from a thermal blanket
Grayscale image on an X-ray film
Fluid flow from an infusion pump
Control Input
Exposure technique settings on an X-ray machine
Sensitivity setting on a medical display
Total infusion volume setting on an infusion pump
Alarm settings on an ECG monitor
 Biomedical Device Technology: Principles and Design

CONSTRAINTS IN BIOMEDICAL SIGNAL MEASUREMENTS

Medical devices in many respects are similar to devices we use in every-

day life. In fact, most technologies used in health care were adapted from the
same technologies used in the military, industrial, and commercial applica-
tions. Since medical devices are used on humans, their reliability and safety
requirements are usually more stringent than other devices. In addition,
medical devices are often used in situations in which patients are vulnerable
to even minor errors; therefore, special consideration in minimizing risk is
necessary in designing medical devices. Listed next are some of the factors
and constraints in designing medical devices.

Low Signal Level

The level of biological signal can be very small, for example, on the
order of microvolts (pV)in EEG measurements. Therefore, very sensitive
transducers as well as good noise rejection methods are required.

Inaccessible Measurement Site

Many signal sources are inside the human body and hidden by other
anatomy. Biomedical measurements and procedures often require invasive
means to obtain access to specific anatomy. For example, to access a nerve
fiber for electrical activity measurement, the electrode must go through the
skin, muscle, and other tissues.

Small Physical Size

Some measurement sites are very small. In order to measure the signal
coming from these tiny sites while at the same time to avoid picking up the
surrounding activities, special sensors that allow isolated measurement at the
source are required. For example, in EMG measurement, needle electrodes
with insulated stems are used to measure the electrical signal produced at a
specific group of muscle fibers.

Difficult to Isolate Signal from Interfering Sources

As we cannot voluntarily turn O N or OFF, or remove tissue or organ to
take a measurement, the measurand is subjected to much interference. As an
example, in fetal monitoring, fetal heart activities are often masked by the
 Overview of Biomedical Instrumentation

stronger maternal heartbeats. It requires special techniques to extract infor-

mation from these interfering signals.

Signal Varies with Time

Human physiological signals are seldom deterministic; they always
change with time and with the activities of the body. It is therefore not an
easy task to establish the norm of such signals. An example of such is in arte-
rial blood pressure measurement; the blood pressure of a person is usually
higher in the morning than at other times of day.

Signal Varies Among Healthy Individuals

Since every human being is different, the same physiological signal from
one is different from that of another. It is not a straightforward task to
establish what is normal or abnormal and what is healthy or unhealthy when
looking at some of these physiological parameters. For example, there is a
huge difference in normal resting heart rate between an athlete and a couch
potato. Nevertheless, there are generally recognized normal ranges. For
example, systolic arterial pressure between 90 and 120 mmHg is considered
acceptable.

Origin and Propagation of Signal Is Not Fully Understood

The human body is very complicated and nonhomogeneous. There are
many signal paths within the body, and interrelationships between physio-
logical events are often not fully understood. For example, the ECG
obtained by surface electrodes looks very different from that obtained by
invasive electrodes placed inside the heart chamber.

Difficult to Establish Safe Level

of Applied Energy to the Tissue
Very often, electrical current from a medical device, whether by inten-
tion or by accident, will flow through the patient's body. Although such ener-
gy will impose risk on the patient, it is often difficult to establish the mini-
mum safety limits of such signals.
 Biomedical Device Technology: Princ$les and Design

Biocompatibility
The parts of a medical device that are in contact with patients must be
nontoxic and must not trigger adverse reaction. In addition, they must be
able to withstand the chemical corrosive environment of the human body.

CONCEPTS ON BIOCOMPATIBILITY

Definitions
Biocornpatibility refers to the compatibility of nonliving materials with
living tissues and organisms, whereas histocompatibility refers to the com-
patibility of different tissues in connection with immunological response.
Histocompatibility is associated primarily with the human lymphocyte anti-
gen system. Rejection of transplants may be prevented by matching tissues
according to histocompatibility and by the use of immunosuppressive drugs.
Biocornpatibility entails mechanical, chemical, pharmacological, and surface
compatibility. It is about the interactions that take place between the materi-
als and the body fluid, tissues, and the physiological responses to these reac-
tions.
Biocornpatibility of metallic materials is controlled by the electrochemi-
cal interaction that results in the release of metal ions or insoluble particles
into the tissues and the toxicity of these released substances. Biocornpatibility
of polymers is, to a large extent, dependent on how the surrounding fluids
extract residual monomers, additives, and degradation products. Other than
the chemistry, biocompatibility is also influenced by other factors such as
mechanical stress imposed on the material.

Mechanism of Reaction
The adverse results of incompatibility include the production of toxic
chemicals, as well as the corrosion and degradation of the biomaterials,
which may affect the function or create failure of the device or implant.
Protein absorption of the implant and tissue infection may lead to premature
failure, resulting in removal and other complications. Compatibility between
medical devices and the human body falls under the heading of biocompat-
ibility.
Biocornpatibility is especially important for implants or devices that for
a considerable length of time are in contact with or inside the human body.
Common implant materials include metal, polymers, ceramics, and products
from other tissues or organisms.
 Ovemiew of Biomedical instrumentation

Tissue Response to Implants

During an implant procedure, the process often requires injuring the tis-
sue. Such injury will invoke reaction such as vasodilation, leakage of fluid
into the extravascular space, and plugging of lymphatics. These reactions
produce classic inflammatory signs such as redness, swelling, and heat, which
often lead to local pain. Soon after injury of the soft tissues, the mesenchy-
ma1 cells evolve into migratory fibroblasts that move into the injured site;
together with the scaffolding formed from fibrinogen in the inflammatory
exudates, collagen is deposited onto the wound. The collagen will dissolve
and redeposit during the next 2 to 4 weeks for its molecules to polymerize in
order to align and create cross-links to return the wound closer to that of nor-
mal tissue. This restructuring process can take more than 6 months.
The body always tries to remove foreign materials. Foreign material may
be extruded from the body (as in the case of a wood splinter), walled off if it
I cannot be moved, or ingested by macrophages if it is in particulate or fluid
1 form. These tissue responses are additional reactions to the healing process-
es described above.
A typical tissue response involves polymorphonuclear leukocytes
appearing near the implant site followed by macrophages (foreign body giant
cells).However, if the implant is chemically and physically inert to the tissue,
only a thin layer of collagenous tissue is developed to encapsulate the
implant. If the implant is either a chemical or a physical irritant to the sur-
rounding tissue, then inflammation occurs at the implant site. The inflam-
mation will delay normal healing and may cause necrosis of tissues by chem-
ical, mechanical, and thermal trauma.
The degree of the tissue response varies according to both the physical
and chemical nature of the implants. Pure metals (except the noble metals)
tend to evoke a severe tissue reaction. Titanium has minimum tissue reaction
' of all the common metals used in implants as long as its oxide layer remains

intact to prevent diffusion of metal ions and oxygen. Corrosion-resistant

alloys such as cobalt-chromium and stainless steel have a similar effect on tis-
sue once they are passivated. Most ceramic materials are oxides such as TiO,
and These materials show minimal tissue reactions with only a thin
layer of encapsulation. Polymers are quite inert toward tissue if there are no
additives such as antioxidants, plasticizers, antidiscoloring agents, et cetera.
On the other hand, monomers can evoke an adverse reaction since they are
very reactive. Therefore, the degree of polymerization is related to the extent
of tissue reaction. As 1000/0polymerization is not achievable, different sizes
of polymer can leach out and cause severe tissue reaction.
A very important requirement for implant or materials in contact with
blood is blood compatibility. Blood compatibility includes creating blood
 Biomedical Device Technology: Principles and Design

clot and damaging protein, enzymes, and blood elements. Damage to blood
elements includes hemolysis (rupture of red blood cells) and triggering of
platelet release. Factors affecting blood compatibility include surface rough-
ness and surface wettability. A nonthrombogenic surface can be created by
coating the surface with heparin, negatively charging the surface, or coating
the surface with nonthrombogenic materials.
Systemic effect can be linked to some biodegradable sutures and surgical
adhesives, as well as particles released from wear and corrosion of metals
and other implants. In addition, there are some concerns about the possible
carcinogenicity of some materials used in implantation.

Characteristics of Materials AEecting Biocompatibility

The following material characteristics should be analyzed to determine
biocompatibility:
Stress-the force exerted per unit area on the material; stress can be
tensile, compressive, or shear.
Strain-the percentage dimensional deformation of the material.
Viscoelesticity-the time-dependent response between stress and
strain.
Thermal properties-include melting point, boiling point, specific heat
capacity, heat capacity, thermal conductivity, and thermal expansion
coefficient.
Surface property-measure of surface tension and contact angle
between liquid and solid surface.
Heat treatment-for example, quenching of metal or surface compres-
sion of glass and ceramics to improve material strength.
Electrical properties-determination of resistivity and piezoelectric
properties.
Optical properties-measurement of refractive index and spectral
absorptivity.
Density and porosity-include measurement of solid volume fraction.
Acoustic properties-include acoustic impedance and attenuation coef-
ficient.
Diffusion properties-determination of permeability coefficients.
Each of these characteristics should be analyzed for its intended applications.
It should be noted that the same material may have different degrees of bio-
compatibility under different environments and different applications.
 Overview of Biomedical Instrumentation

In Vitro and In Vivo Tests for Biocompatibility

Tests for biocompatibility should include both material and host respons-
es. The usual approach in testing a new product is to perform an in vitro
screening test for quick rejection of incompatible materials. In vitro tests can
be divided into two general classes: (1) tissue culture methods, and (2) blood
contact methods.
Tissue culture refers collectively to the practice of maintaining portions
of living tissues in a viable state, including cell culture, tissue culture, and
organ culture. Blood contact methods are performed only for blood contact
applications such as cardiovascular devices. Both static and dynamic (flow)
tests should be performed.
After screening by in vitro techniques, the product is moved to in vivo
testing. It is the practice to test new implant materials or existing materials in
significantly different applications. In vivo tests are often done in extended-
time whole animal tests before human clinical trials. In vivo tests in general
are divided into two types: (1) nonfunctional, and (2) functional.
In nonfunctional tests, the product material can be of any shape and is
embedded passively in the tissue site for a period of time (e.g., a few weeks
to 24 months). Nonfunctional tests focus on the direct interactions between
the material of the product and the chemical and biological species of the
implant environment. In addition to being implanted, functional tests require
that the product be placed in the functional mode as close as possible to the
conditions of its intended applications. The purpose of functional tests is to
study both the host and material responses such as tissue in growth into
porous materials, material fatigue, and production of wear particles in load-
bearing devices. Functional tests are obviously more involved and costly
than nonfunctional tests.

FUNCTIONAL BUILDING BLOCKS

OF MEDICAL INSTRUMENTATION

A typical diagnostic medical device acquires information from the

patient, analyzes and processes the data, and presents the information to the
clinician. In a therapeutic device, it processes the input from the clinician
and applies the therapeutic energy to the patient. Figure 1-13 shows the
functional building blocks of a typical medical device. It includes the follow-
ing functional building blocks:
Patient interface
Analog processing
 Biomedical Device Technology: Princ$les and Design

Analog to digital conversion and digital to analog conversion

Signal isolation
Digital processing
Memory
User interface

Digital
Conversion

Digital
Interface Processing Isolation Processing Interface

Digital to
Analog
tl Memory

Figure 1-13. Functional Block Diagram of a Medical Instrument.

Patient Interface
In diagnostic devices, the patient interface includes transducers or sen-
sors to pick up and convert the physiological signal (e.g., blood pressure) to
an electrical signal. In therapeutic devices, the patient interface contains
transducers that generate and apply energy to the patient (e.g., ultrasound
physiotherapy unit).

Analog Processing
The analog processing contains electrical circuits such as amplifiers (to
increase signal level) and filters (to remove any unwanted frequency compo-
nents such as high-frequency noise from the signal). The signal until this
point is still in its analog format.

Analog to Digital Conversion

The function of the analog to digital converter (ADC) is to convert the
analog signal to its digital format. The signal coming from the ADC is a
string of binary numbers (1's and 0's).
 Overview of Biomedical Instrumentation

Digital to Analog Conversion

If an analog output is necessary, a digital to analog converter (DAC) will
be required to convert the digital signal from its "I" and "0" states back to its
analog format. A DAC reverses the process of an ADC.

Signal Isolation
The primary function of signal isolation is for microshock prevention in
patient electrical safety. The isolation barrier, usually an optocoupler, pro-
vides a very high electrical impedance between the patient's applied parts
and the power supply circuit to limit the amount of risk current flowing to or
from the patient.

Digital Processing
After being digitized by the ADC, the signal is sent to the digital pro-
cessing circuit. In a modern medical instrument, digital processing is done by
one or more computers built into the system. The center of a digital com-
puter is the central processing unit (CPU).Depending on the needs, the CPU
may perform functions such as calculations, signal conditioning, pattern
recognition, information extraction, et cetera.

Memory
Information such as waveforms or computed data is stored in its binary
format in the memory module of the device. Signal stored in the memory
later be retrieved for display, analysis, or used to control other outputs.

User Interface
User interfaces can be output or input devices. Examples of output user
interfaces are video displays for physiological waveforms and audio alarms.
Examples of input devices are touch screen and trackballs.
>
-
 Chapter 2

CONCEPTS IN SIGNAL MEASUREMENT,

PROCESSING, AND ANALYSIS

OBJECTIVES

List the typical elements in medical device specifications.

Define common parameters found in device specifications, including
accuracy, error, precision, resolution, reproducibility, sensitivity, lin-
earity, hysteresis, and zero offset.
Describe common steady state nonlinear input-output characteristics.
Define transfer functions in time and frequency domains.
Explain and analyze the effect of filters on biomedical signals.

CHAFTER CONTENTS

1. Introduction
2. Device Specifications
3. Steady State Versus Transient Characteristics
4. Linear Versus Nonlinear Steady State Characteristics
5. Time and Frequency Domains
6. Signal Processing and Analysis

INTRODUCTION

Most medical devices involve measurement or sensing one or more

physiological signals, enhancing the signals of interest, and extracting useful
 Concepts in Signal Measurement, Processing, and Analysis

information from the signals. The concepts of signal measurement, process-

ing, and analysis are fundamental in understanding scientific instrumenta-
tions, including medical devices. This chapter provides an introductory
overview of these concepts.

DEVICE SPECIFICATIONS

To understand the functions and performance of a medical device, one

should start from reading the specifications of the device. Specifications of an
instrument are the claims from the manufacturer on the performance of the
instrument. The specification document of a medical device should contain
at least the following information:
List of device functions
Input and output characteristics
Performance statements
Physical characteristics
Environmental requirements
Codes and regulations
Below is an example of a specification document of an electrocardio-
graph:

Instrument type: 12-channel, microcomputer-augmented, automatic electrocardiograph

Input channels: simultaneous acquisition of up to 12 channels
Frequency response: -3 dB @ 0.01 to 105 Hz
Sensitivities: 2.5, 5, 10, and 20 mrn/mV, k20/0
Input impedance: > 50 MO
Common mode rejection ratio: > 106 dB
Recorder type: thermal digital dot array, 200 dots/in vertical resolution
Recorder speed: 1,5, 25, and 50 rnm/sec, +3"h
Digital sampling rate: 2,000 samples/sec/channel
ECG analysis frequency: 250 sarnples/sec
Display formats: user-selectable channel and lead configurations: 3, 4, 5, 6, and 12
channels
Dimensions: H X W X D = 90 cm X 42 cm X 75 cm
Weight: 30 kg
Power requirements: 90 VAC to 260 VAC, 50 or 60 Hz
Certification: UL 544 listed, meets ANSIIAAMI standards, complies with IEC 601
standards
 Biomedical Device Technology: Principles and Design

Some common parameters found in medical device specifications are

explained next.
The error ( E ) of a single measured quantity is the measured value ( Q m )
minus the true value (Qt), or
E = Qm - Qt.

There are three types of errors: gross error, systematic error, and random
error.
Gross error arises from incorrect use of the instrument (e.g., human
error).
Systematic error is due to a shortcoming of the instrument (e.g., defec-
tive or worn parts, adverse effect of the environment on the equipment).
Random error is fluctuations that cannot be directly established or cor-
rected (e.g., noise in photographic process).
Error may be expressed as absolute or relative.
Absolute error is expressed in the specific units of measurement, e.g.,
+
15 0 1 0. The graphical representation is shown in Figure 2-la.
Relative error is expressed as a ratio of the measured quantity, e.g.,
output reading + 5%. (Figure 2-lb)
An alternative way to express absolute error is percentage of full scale,
e.g., 5010 of full scale output. (Figure 2-lc).
+
Or it can be a combination of the above, e.g., 1 Cl or 5% of output,
whichever is greater. (Figure 2- 1d)
Accuracy (A) is the error divided by the true value and is often
expressed as a percentage.

Accuracy usually varies over the normal range of the quantity measured. It
can be expressed as a percentage of the reading or a percentage of full scale.
For example, for a speedometer with *5.0% accuracy, when it is reading 50
km/hr, the maximum error is +2.5 km/hr. If the speedometer is rated at
?5.O0Io full scale accuracy and the full scale reading is 200 km/hr, the maxi-
mum error of the measurement is ? 10 km/hr, irrespective of the reading.
The precision of a measurement expresses the number of distinguish-
able alternatives from which a given result is selected. For example, a meter
that can measure a reading of three decimal places (e.g., 4.123 V) is more
precise than one than can measure only two decimal places (e.g., 4.12 V).
Resolution is the smallest incremental quantity than can be measured
with certainty. If the readout of a digital thermometer jumped from 20°C to
22°C and then to 24°C when it is used to measure the temperature of a bath
 Concepts in Sigrzral Measurement, Processing, and Analysis

Output Output

+Input Input
a) Absolute Error b) Percentage Error

Output Output

)Input + Input
Ic) Percentage Full Scale Error I d) Combination of Absolute and
Percentage Error
Figure 2-1. Error Representation.

of water slowly being heated by an electric water heater, the resolution of the
thermometer is 2'C.
Reproducibility is the ability of an instrument to give the same output
for equal inputs applied over some period of time.
Sensitivity is the ratio of the incremental output quantity to the incre-
mental input quantity (S = 2).
It is the slope or tangent of the output versus
input curve. Note that the sensitivity of an instrument is a constant only if the
output-input relationship is linear. For a nonlinear transfer function (as
shown in Figure 2-2), the sensitivity is different at different points on the
curve (Si # S2).
Zero offset is the output quantity measured when the input is zero. Input
zero offset is the input value applied to obtain a zero output reading. Zero
offsets can be positive or negative.
Zero drift has occurred when all output values increase or decrease by
the same amount.
A sensitivity drift has occurred when the slope (sensitivity) of the input-
output curve has changed over a period of time.
Perfect linearity of an instrument requires that the calibration curve be
 Biomedical Device Technology: Principles and Design

Figure 2-2. Sensitivity.

a straight line. That is, a linear instrument has the characteristics

where x is the input, y is the output, and m and c are both constants.
Independent nonlinearity expresses the maximum deviation of points
from the least-squares fitted line as either f POI0 of the reading or ? p/o of full
scale, whichever is greater. Percentage nonlinearity (Figure 2-3) is defined as
the maximum deviation of the input (Dm) from the curve to the least square
fit straight line divided by the full scale input range (I#).It is sometimes referred
to as 010 input nonlinearity (versus OIo output nonlinearity).

010 nonlinearity = -x 100%.

Ifs

Output

-----.-.-
Experiment
Data
Least Square
Fit Straight Line

I I

I I
b Input
I + - - - - - - - - - - - - - - - ( --------------,I
fs

Figure 2-3. Percentage Nonlinearity.

 Concepts in Signal Measurement, Processing and Analysis

An instrument complies with the listed specifications (such as accuracy,

010 nonlinearity) only within the specified input ranges. In other words,
when the input of the instrument is beyond the specified input range, its
characteristics may not be according to the labeled specifications.
Since biomedical transducers and instruments usually convert nonelec-
trical quantities to electrical quantities (voltage or current), its input imped-
ance must be specified to evaluate the degree to which the instrument dis-
turbs the quantity being measured.
Hysteresis measures the capability of the output to follow the change of
the input in either direction. Hysteresis often occurs when the process is
lossy.
Response time is the time required for the output to change from its
previous state to a final settled value given the tolerance. (e.g., time to change
to 90% of the steady state).
Calibration is the process of determining and recording the relation-
ships between the values indicated by a measuring instrument and the true
value of the measured quantity. Since the true value is usually difficult to
obtain, the instrument is usually calibrated against a device that is traceable
to a national standard.
Statistical control ensures that random variations in measured quanti-
ties (result from all factors that influence the measurement process) are tol-
erable. If random variables make the outputs nonreproducible, statistical
analysis must be used to determine the error variation. In fact, many med-
ical devices rely on statistical means to determine their calibration accuracy.

STEADY STATE VERSUS TRANSIENT CHARACTERISTICS

For a typical instrument, the output will change following a change in the
input. Figure 2-4 shows a typical output response when a step input is
applied to the system. Depending on the system characteristics, the output
may experience a delay before it settles down (dotted line in Figure 2-4) or
may get into oscillation right after the change of the input (solid line in Figure
2-4). However, in most instruments, this transient will eventually settle down
to a steady state until the input is changed again.
The input-output characteristics when one ignores the initial transient
period is called the steady state characteristics or static response of the sys-
tem. When the input is a time-varying signal, one must take into account the
transient characteristics of the system. For example, when the input is a fast-
changing signal, the output may not be able to follow the input, that is, the
output may not have enough time to reach its steady state before the input is
changed again. In this case, the signal will suffer from distortion.
 Concepts in S i p a l Measurement, Processing, and Analysis

An instrument complies with the listed specifications (such as accuracy,

oh nonlinearity) only within the specified input ranges. In other words,
when the input of the instrument is beyond the specified input range, its
characteristics may not be according to the labeled specifications.
Since biomedical transducers and instruments usually convert nonelec-
trical quantities to electrical quantities (voltage or current), its input imped-
ance must be specified to evaluate the degree to which the instrument dis-
turbs the quantity being measured.
Hysteresis measures the capability of the output to follow the change of
the input in either direction. Hysteresis often occurs when the process is
lossy.
Response time is the time required for the output to change from its
previous state to a final settled value given the tolerance. (e.g., time to change
to 90% of the steady state).
Calibration is the process of determining and recording the relation-
ships between the values indicated by a measuring instrument and the true
value of the measured quantity. Since the true value is usually difficult to
obtain, the instrument is usually calibrated against a device that is traceable
to a national standard.
Statistical control ensures that random variations in measured quanti-
ties (result from all factors that influence the measurement process) are tol-
erable. If random variables make the outputs nonreproducible, statistical
analysis must be used to determine the error variation. In fact, many med-
ical devices rely on statistical means to determine their calibration accuracy.

STEADY STATE VERSUS TRANSIENT CHARACTERISTICS

For a typical instrument, the output will change following a change in the
input. Figure 2-4 shows a typical output response when a step input is
applied to the system. Depending on the system characteristics, the output
may experience a delay before it settles down (dotted line in Figure 2-4) or
may get into oscillation right after the change of the input (solid line in Figure
2-4). However, in most instruments, this transient will eventually settle down
to a steady state until the input is changed again.
The input-output characteristics when one ignores the initial transient
period is called the steady state characteristics or static response of the sys-
tem. When the input is a time-varying signal, one must take into account the
transient characteristics of the system. For example, when the input is a fast-
changing signal, the output may not be able to follow the input, that is, the
output may not have enough time to reach its steady state before the input is
changed again. In this case, the signal will suffer from distortion.
 Biomedical Deuice Technology: Binc$les and Design

Output
4

Time
Figure 2-4. Output Response to a Step Input.

LINEAR VERSUS NONLINEAR STEADY

STATE CHARACTERISTICS

A linear calibration curve is one that obeys the relationship

y=mx+c
where x is the input,
y is the output, and
m and c are both constants.
Figure 2-5a shows a linear characteristics with c = 0.
Some common nonlinear characteristics are shown in Figure 2-5.
Saturation occurs when the input is increased to a point where the
output cannot be increased further. A linear operational amplifier will
become saturated when the input is close to the power supply voltage.
Breakdown is the phenomenon when the output abruptly starts to
increase when the input changes slightly following a linear relation-
ship. Some devices such as Zener diodes have this type of nonlinear
behavior.
Dead zone is a range of the input where the output remains constant.
A worn-out gear system usually has some dead space (dead zone).
Bang-bang occurs when a minor reversal of the input creates an
abrupt change in the output. This phenomenon can be observed in
some thin metal diaphragm transducers. The diaphragm may flip from
one side to another when the force applied to the center of the
diaphragm changes direction.
Hysteresis is the phenomenon that the input-output characteristic fol-
 Concepts in Sipal Measurement, Processing, and Analysis

lows different pathways depending on whether the input is increasing

or decreasing. Hysteresis results when some of the energy applied dur-
ing an increasing input is not recovered when the input is reversed.
The magnetization characteristic of a transformer is a perfect example
where some applied energy is loss in the eddy current in the iron core.

Output Output

a) Linear b) Saturation

Output Output
C C

b
Input

c) Breakdown d) Dead Zone

Output Output

e) Bang-bang f) Hysteresis
Figure 2-5 a) to f). Common Nonlinear Characteristics.
 Biomedical Device Technology: Principles and Design

TIME AND FREQUENCY DOMAINS

A time-varying signal is a signal the amplitude of which changes with

time. A periodical signal is a time-varying signal the waveshape of which
repeats at regular time intervals. Mathematically, a periodical signal is given
by:

where n = any positive or negative integer,

and T = fundamental period.
A sinusoidal signal is an example of a periodical signal. The mathemati-
cal expression of a sinusoidal signal is

where A = a constant,
w = angular velocity, and
+ = phase angle.
Any periodical signal can be represented (through Fourier-series expan-
sion) by a combination of sinusoidal signals

where ao, a? and bn are time-invariant values depending on the shape of

G(t), and o 0 = 27Tf0 = 27~/T.
Using the preceding equation, the Fourier series of a symmetrical square
+
waveform (Figure 2-6a) with amplitude of 1 V and a period of 1 sec is:
V(t) = 4 V / r (sin 2nt + 0.33sin6.rrt + 0.20sinlOrt + 0.14sinl4.rrt +...).
The frequency domain plot or frequency spectrum of the same signal is
shown in Figure 2-6b. A brief description of Fourier analysis can be found
in Appendix A.1.
For a nonperiodical signal, the frequency spectrum is continuous instead
of discrete. An example of an arterial blood pressure waveform is shown in
Figure 2-7.
All biomedical signals are time-varying. Some signals may change very
slowly with time (e.g, body temperature), while others may change more rapid-
ly (e.g., an ECG). Although some appear to be periodical, in fact, each cycle of
the signal differs from the others due to many factors. Physiological signals that
appear to be periodical are considered pseudoperiodical. For example, each
cycle of the invasive blood pressure waveform of a resting healthy person may
look the same within a short period of time; however, the waveform and
 Concepts in S i p a l Measurement, Processing, and Analysis

Amplitude (V)
C

Figure 2-Ga. Square Waveform in Time-Domain.

Amplitude (V)

t A = 417~

I Frequency (Hz)
Figure 2-Gb. Frequency Spectrum of the Square Wave in a.

Pressure (mmHg) Magnitude

I I I I I b
0 0.4 0.8 Time (s) 0 4 6 8 10 Frequency (Hz)
Time Domain Frequency Domain
Figure 2-7. Arterial Blood Pressure Signal.

amplitude will be quite different when the person is engaged in different phys-
ical activities such as running. Furthermore, the waveform may be very differ-
ent from cycle to cycle when the person has cardiovascular problems.
 Biomedical Device Technology: Princ$les and Design

SIGNAL PROCESSING AND ANALYSIS

The purpose of signal processing and analysis in medical instrumentation

is to extract useful information from the "raw" biological signals. For exam-
ple, an ECG monitor can derive the rate of the heartbeats from the biopo-
tential waveform of the heart activities. It also can generate an alarm signal
to alert the clinician should the heart rate fall outside a predetermined range
(e.g., greater than 120 bpm or less than 50 bpm).

Transfer Function
Mathematically, an operation or process (Figure 2-8) can be represented
by a transfer function f(t). When a signal x(t) is processed by the transfer
function, the output y(t) is equals to the time convolution between the input
signal and the transfer function. That is,
y(t) = f(t -h)x(h)dh
or simply denoted by y(t) = f(t)* x(t).

Figure 2-8. Time Domain Transfer Function.

In the frequency domain, the mathematical relationship between the out-

put and input signals when the input is processed by the transfer function F
(Figure 2-9) is given by:
Y(w) = F(w) X(w),
where X(w) is the input signal,
Y(w) is the output signal,
F(w) is the transfer function of the process, and
w = angular velocity = 2nf.
Note that the output is simply equal to the input multiplied by the trans-
fer function in the frequency domain. This is why signal analysis is often per-
formed in the frequency domain.
 Concepts in Signal Measurement, Processing, and Analysis

Figure 2-9. Frequency Domain Transfer Function.

Signal Filtering
A filter separates signals according to their frequencies. Most filters
accomplish this by attenuating the part of the signal that is in one or more
frequency regions. The transfer function of a filter is frequency-dependent. A
filter can be represented by a transfer function F(o). Filters can be low pass,
high pass, band pass, or band reject. The four types of filters are shown in
Figure 2-10. The cutoff frequency (corner frequency) of a filter is usually
measured at -3dB from the midband amplitude (70.7% of the amplitude).
A low pass filter attenuates high frequencies above its cutoff frequency.
An example of such is the filter used to remove baseline wandering signal in
ECG monitoring; a 0.5Hz high pass filter is switched into the signal path to
remove the low-frequency component caused by the movement of the
patient. High pass filters attenuate low frequencies and allow high-frequency
signals to pass through. Many biomedical devices have low pass filters with
upper cutoff frequencies to remove unwanted high-frequency noise. A band
pass filter is a combination of a high pass filter and a low pass filter, it elimi-
nates unwanted low- and high-frequency signals while allowing the mid-fre-
quency signals to go through. A band reject filter removes only a small band-
width of frequency signal. A 60 Hz notch filter designed to remove 60 Hz
power-induced noise is an example of a band reject filter. Filters can be
inserted at any point in the signal pathway. Filters can be inherent (charac-
teristics of the intrinsic or parasitic circuit components) or inserted to achieve
a specific effect. For example, a low pass filter is inserted in the signal path-
way to remove high-frequency noise from the signal, which results in a
"cleaner" waveform.
Figure 2-11 shows the effect of filters on an ECG waveform. Figure
2-lla is acquired using a bandwidth from 0.05 to 125 Hz. In Figure 2-llb,
the upper cutoff frequency is reduced from 125 Hz to 25 Hz. The effect of
eliminating the high-frequency components in the waveform is the attenua-
tion of the fast-changing events (i.e., reduction of the amplitude of the R-
wave). Figure 2-llc shows the effect of increasing the lower cutoff frequen-
cy from 0.05 to 1.0 Hz. In this case, the low-frequency component of the sig-
nal is removed. Therefore, the waveform becomes more oscillatory.
 Biomedical Device Technology: Princ$les and Design

Normalized Normalized
Amplitude Amplitude
A

1.0 , 1.o

b b
Frequency (f) Frequency (f)
a) Low Pass a) High Pass

Normalized Normalized
Amplitude Amplitude

1.o 1.0 ,

b b
Frequency (f) Frequency (f)
c) Band Pass d) Band Reject
Figure 2-10. Filter Transfer Functions.

Signal Amplification and Attenuation

An amplifier increases (amplifies) the signal amplitude, while an attenu-
ator decreases (attenuates) the signal amplitude. The transfer function of an
amplifier is also called the amplification factor (A). A is expressed as the ratio
of the output (Y) to the input quantity (X). That is:

The amplification factor can be frequency-dependent (i.e., the value of A is

different at different frequencies). The transfer function of an ideal amplifier
or attenuator is independent of time and frequency (i.e., it has a constant
magnitude at all frequencies).
 Concepts in Signal Measurement, Processing, and Analysis

1 - 1 1 5 ff*

Figure 2-11. Effect of Filters on ECG Waveform.

Other Signal Processing Circuits

Other than filters and amplifiers, there are many other signal processors
with different transfer function characteristics. Examples are integrators, dif-
ferentiators, multipliers, adders, inverters, comparators, logarithm amplifiers,
et cetera. Readers who want to learn more about these signal processing cir-
cuits should refer to an integrated electronic textbook. Although many med-
ical devices still use analog signal processing circuits, more and more of these
signal processing functions are performed digitally by software in modern
computer-based devices.
 Chapter 3

FUNDAMENTALS OF BIOMEDICAL
TRANSDUCERS

OBJECTIVES

Define the terms transducer, sensor, electrode, and actuator.

Distinguish between the following modes of biological signal measure-
ments: direct and indirect, intermittent and continuous, desired and
interfering, invasive and noninvasive.
Specify the three criteria for the faithful reproduction of an event.
Evaluate the effect on physiological signal measurements due to arnpli-
tude nonlinearity, phase distortion, and bandwidth limitation.
Analyze Wheatstone bridge circuits in medical instrumentation appli-
cations.

CHAPTER CONTENTS

1. Introduction
2. Definitions
3. Types of Transducers
4. Transducer Characteristics
5. Signal Conditioning
6. Transducer Excitation
7. Common Physiological Signal Transducers
 Fundamentals of Biomedical Transducers

INTRODUCTION

Medical devices are designed either to measure physiological parameters

from the patient or to apply certain energy to the patient. To achieve that, a
medical device must use a transducer to interface between the device and the
patient. The transducer is usually the most critical component in a medical
device as it must reliably and faithfully reproduce the signal to or from the
patient. In addition, transducers are often in contact with or even implanted
inside a human. Transducers in such applications must be stable and non-
toxic to the human body. Ideally, a transducer should respond only to the
energy that is desired to be measured and exclude the others. Most of the
medical device constraints discussed in Chapter 1 are also applicable to bio-
medical transducers.

DEFINITIONS

A sensor is a device that can sense changes of one physical quantity and
transpose them systematically into a different physical quantity. Generally
speaking, a transducer is defined as a device to convert energy from one
form to another. For example, the heating element on the kitchen stove is a
transducer that converts electrical energy to heat energy for cooking. In
instrumentation, a transducer is a device whose main function is to convert
the measurand to a signal that is compatible with a measurement or control
system. The compatible signal is often an electrical signal. For example, an
optical transducer may convert light intensity to an electrical voltage. In
instrumentation or measurement applications, sensors and transducers are
often use synonymously. An electrode is a transducer that directly acquires
the electrical signal without the need to convert it to another form; that is,
both input and output are electrical signals. On the other hand, an actuator
is a transducer that produces a force or motion. An electrical motor is an
example of an actuator that converts electricity to mechanical motion.
In many biomedical applications, the transducer or sensor converts a
physiological event to an electrical signal. With the event available as an
electrical signal, it is easier to use modern computer technology to process
the physiological event and display the output in a user-friendly format.
Figure 3-1 shows a simple block diagram of a physiological monitor.
 Biomedical Device techno lo^: Pfinc$les and Design

Patient ---b Transducer + Conditioning

Signal
+
, Display .--+Clinician
Figure 3-1. Physiological Monitor.

Passive Versus Active Transducers

Transducers can be passive or active. For a passive transducer, the input
to the transducer produces change in a passive network parameter such as
resistance, capacitance, or inductance. On the other hand, an active trans-
ducer, such as a piezoelectric crystal or a thermocouple, acts as a generator,
producing force, current, or voltage in response to the input.

Direct Versus Indirect Mode of Transducers

For a direct transducer, the measurand is interfaced to and measured
directly by the transducer. Blood pressure may be measured directly by plac-
ing a pressure transducer inside a blood vessel. The electrical cardiac signal
is directly picked up by a set of electrodes placed on the chest of a patient.
Both are examples of direct mode of transduction. In indirect mode, the
transducer measures another measurand that bears a known relationship to
the desired measurand. Indirect transducers are often used when the desired
measurands are not readily accessible. An example of such indirect mode of
transduction approach is in noninvasive blood pressure measurements,
where the systolic, mean, and diastolic pressures are estimated from the oscil-
latory characteristics of the pressure in a pneumatic cuff applied over the
upper arm of the subject.

Intermittent Versus Continuous Measurement

In some cases, it is important to monitor physiological parameters in a
continuous manner. Continuously monitoring the heart rate and blood oxy-
gen level of a patient during general anesthesia is an example. In other cir-
cumstances, a single measurement is sufficient to obtain a snapshot of the
patient's condition. Measurement of oral temperature using a liquid-in-glass
thermometer is an example of intermittent temperature measurement. In
another application, periodical measurement to track changes is more appro-
priate. Charting the arterial blood pressure of a patient in the recovery room
every 5 minutes is an example of periodical measurement.
 Fundamentals of Biomedical Transducers

Desired Versus Interfering Input

Desired input to a transducer is the signal that the transducer is designed
to pick up. Interfering input is an unwanted signal that affects or corrupts the
output of the transducer. For example, maternal heart rate is the interfering
input in fetal heart rate monitoring. Interfering input is sometimes referred
to as noise in the system. In medical applications, interfering input is usual-
ly compensated for by adjusting the sensor location, or through signal pro-
cessing such as filtering.

Invasive Versus Noninvasive Method

A procedure that requires bypassing the skin of the patient is called an
invasive procedure. Entering the body cavity such as through the mouth into
the trachea is also considered to be invasive. In biomedical applications,
measurement of a physiological signal often requires placing a transducer
inside the patient's body. Using a needle electrode to measure myoelectric
potential is an example of an invasive method of measurement. On the other
hand, myoelectric measurements using skin (or surface) electrodes are non-
invasive procedures.

TRANSDUCER CHARACTERISTICS

A transducer is often specified by the following:

The quantity to be measured, or the measurand
The principle of the conversion process
The performance characteristics
The physical characteristics
In biomedical measurements, common measurands are position, motion,
velocity, acceleration, force, pressure, volume, flow, heat, temperature,
humidity, light intensity, sound level, chemical composition, electric current,
electric voltage, et cetera. Examples of their characteristics and method of
measurements of some of these physiological parameters are tabulated in
Table 1-2 in Chapter 1.
Many methods can be used to convert a physiological event to an elec-
trical signal. The event can be made to modify, directly or indirectly, the
electrical properties of the transducer, such as its resistance or inductance
values. The primary functional component of a transducer or sensor is the
transduction element. Many different transduction elements are suitable for
health care applications. Table 3-1 lists some common transducer categories
 Biomedical Device Technology: Princ$les and Design

and their operating principles. Examples of transducers in each category are

also listed in the table.

Table 3-1.
'hansducers and Their Operating Principles
- - - -

Transduction Correlating Examples

-
Elements
- - - -
Properties
- -

Resistive Resistance-temperature Thermistor to measure temperature

Resistance-displacement Resistive strain gauge to measure
due to pressure pressure
Capacitive Capacitive blanket in neonatal apnea
detection
Inductive Inductance-displacement Linear variable differential transformer
due to pressure in pressure measurement
Photoelectric Electric current- light Photomultiplier in scintillation counter,
energy red and infrared LEDs and detectors
in pulse oximeter
Piezoelectric Electric potential-force Ultrasound transducer in blood flow
detector
Thermoelectric Electric potential- thermal Thermocouple junctions in temperature
energy measurement
Chemical Electric current-chemical Polarographic cell in oxygen analyzer
concentration
- -- --

Transducers should adhere to the following three criteria for the faithful
reproduction of an event:
1. Amplitude linearity-ability to produce an output signal such that its
amplitude is directly proportional to the input amplitude.
2. Adequate frequency response-ability to follow both rapid and slow
changes.
3. Free from phase distortion-ability to maintain the time differences in
the sinusoidal frequencies.

Amplitude Linearity
The output and input should follow a linear relationship within its oper-
ating range. The output will not resemble the input if the above is not true.
A common example of a nonlinear input-output relationship is saturation of
operational amplifier when the input becomes too large. In Figure 3-2, when
the input is within the linear region of the operational amplifier, any change
of the input will produce a change of output proportional to the change of
 Fundamentals of Biomedical Transducers

input. When the input becomes too large, it drives the amplifier into satura-
tion with the effect that the output will not increase further with the input;
the waveform is "clipped."

9 r
Input

Figure 3-2. Effect of Nonlinearity (Saturation).

Adequate Frequency Response

In physiological signal measurements, the signals often change with time;
body temperature changes slowly, whereas the heart potential (ECG)
changes more rapidly. In order to accurately measure a changing signal, the
transducer should be able to follow the changes of the input; that is, it must
have a wide enough frequency response. Figure 3-3 shows the effect of inad-
equate frequency response. In this case, the high frequency of the ECG sig-
nal is attenuated or removed by the low pass filtering effect of the system.
Note that in the output waveform, the amplitude of the R-wave is substan-
tially reduced.

lnput

Figure 3-3. Effect of Inadequate High Frequency.

 Biomedical Device Technology: Principles and Design

Free from Phase Distortion

A system that creates different time delay at different signal frequency
will create phase distortion. As we know, any time-varying signal can be rep-
resented by a number of sinusoidal signals of different frequencies and
amplitudes, and recombining or adding these signals will reproduce the orig-
inal signal (Fourier analysis). However, if the transducer in the measurement
process creates different time delays on the sinusoidal signal components,
recombining these signals at the output of the transducer will produce a dis-
torted signal (Figure 3-4). Phase distortion will prevent the faithful repro-
duction of an event.

Sd---$l--- Input

Figure 3-4. Effect of Phase Delay.

Any deviations from these three criteria will produce distorted output
signals. Therefore, transducers must be carefully chosen to minimize distor-
tion within the range of measurement. If signal distortion cannot be avoided
due to nonideal transducers, additional electronic circuits may be used to
compensate for such distortions.

SIGNAL CONDITIONING

A transducer output may be directly coupled to a display device to be

viewed by the user. Very often, the output of a transducer is coupled to a sig-
nal conditioning circuit. A signal conditioning circuit can be as simple as a
passive filter or as complicated as a digital signal processor. A very common
signal conditioning circuit for passive transducers is a Wheatstone bridge.
Many functional elements for signal conditioning commonly used in indus-
trial electrical instrumentations are used in medical devices. These function-
 Fundamentals ofBiomedical Transducers

a1 elements can be implemented using analog components or performed by

software. In the latter case, the signal must be digitized and processed by a
digital computer. Some common signal conditioning functions are amplifiers,
filters, rectifiers, peak detectors, differentiators, integrators, et cetera.

TRANSDUCER EXCITATION

Transducers that vary their electrical values according to changes in their

inputs are often used in biomedical applications. These transducers are usu-
ally coupled to operational amplifiers to increase their sensitivities and to
reject noise. When the output of a transducer is a passive electrical parame-
ter (e.g., resistance, capacitance, or inductance), it often requires an excita-
tion to convert the passive output variable to a voltage signal. The excitation
can be a constant voltage or a constant current source; it may be a DC or an
AC signal of any frequencies. A common excitation method in biomedical
transducers is the Wheatstone bridge.
A Wheatstone bridge is commonly used to couple a transducer to the
other electronic circuits. Figure 3-5 shows a typical Wheatstone bridge with
excitation voltage VE and impedances 21, 22, 23, and Z4 at each arm of the
bridge.
The output of the bridge Vo in the figure is:

From the equation, when Z2

) = 0, the bridge output

voltage is zero (Vo = 0).

z
4
VE
+ ""- b

Figure 3-5. Wheatstone Bridge Circuit.

 Biomedical Device Technology: Princ$les and Design

In most transducer applications, one of the bridge arm impedances is

replaced by a transducer whose impedance changes with the parameter
being measured. Figure 3-6 shows an example of such an arrangement. The
+
transducer impedance can be written as Z AZ, where AZis the impedance
that changes with the quantity being measured and Zis the invariable part of
the transducer impedance. In this example, the impedances in the remaining
arms are all equal to Z

Figure 3-6. Wheatstone Bridge Transducer Circuit.

Substituting ZI = Z + AZ, and Zz = 2 3 = 24 = Z i n equation (1)gives:

Equation (2) shows that Vo and AZ has a nonlinear relationship. How-

ever, if AZis much smaller than Z (AZ << Z), equation (2) can be approxi-
mated by:

Now consider the half bridge circuit in Figure 3-7, where there are two
transducers Z + AZand Z - AZ, each on one arm of the bridge.
+
Substituting the Zr = Z AZ, 2 2 = Z - A& 23 = 2 4 = Zin equation (1)
gives:

Note that the bridge output Vo is proportional to the change in transducer

impedance AZwith a proportionality constant (or sensitivity) equals to -
22
 Fundamentals of Biomedical Transducers

If we replace all the fixed impedances on the bridge arms with transduc-
ers as shown in Figure 3-8, the bridge circuit is called a full bridge.

Z-AZ

Figure 3-8. Full Bridge Transducer Circuit.

Substituting the 21= 23 = = Z + AZ, 2 2 = 2 4 = Z - AZ in equation (1)

gives:

Similar to the half bridge, the full bridge output Vo is proportional to the
change in transducer impedance AZ. However, the proportionality constant
VE , which is two times the value of a half bridge
for a full bridge circuit is -
z
circuit.
Among the three bridge circuits discussed, a full bridge transducer circuit
produces a linear output voltage with respect to changes in the transducer
 Biomedical Device Technology: Principles and Design

impedance and it has the highest sensitivity. However, it requires four match-
ing transducers compared to two for the half bridge and only one for the typ-
ical bridge circuit.

COMMON PHYSIOLOGICAL SIGNAL TRANSDUCERS

In biomedical measurement, the transducer is a component of the med-

ical device that picks up the physiological signal from the patient. It is also
the interface between the device and the human body. The transducer with
its excitation circuit and its associated analog signal processing components
is sometimes referred to as the frontend of the medical device. There are
many types of transducers in biomedical applications. Their characteristics,
principles of operation, and design can be very different. The following
major categories of biomedical transducers are covered in the next six chap-
ters:
1. Pressure transducers
2. Temperature transducers
3. Motion transducers
4. Flow transducers
5. Optical transducers
6. Electrochemical transducers
 Chapter 4

PRESSURE AND FORCE TRANSDUCERS

Understand different units of pressure measurement and perform unit

conversion.
Define absolute and gauge pressure.
Explain the function of barometers and manometers.
State the principles of bourdon tube, bellow, and diaphragm pressure
meters.
Define gauge factor for a resistive stain gauge and derive its dependent
variables.
Describe bonded and unbonded strain gauges, metal strain wire, and
diaphragm strain gauges.
Analyze the principles of piezoelectric pressure gauges.

CHAPTER CONTENTS

1. Introduction
2. Barometers and Manometers
3. Mechanical Pressure Gauges
4. Strain Gauges
5. Piezoelectric Pressure Transducers
 Biomedical Device Technology: Principles and Design

INTRODUCTION

If a force F i s acting uniformly on and perpendicular to a surface of area

A, the pressure P on the surface is defined as:

From this definition, a force transducer may be used for pressure measure-
ment and vice versa. Pressure transducers have many applications in bio-
medical instrumentation. Blood pressure measurement is one of the routine
procedures performed in medicine. Several types of pressure-sensing ele-
ments are discussed in this chapter.

BAROMETERS AND MANOMETERS

One of the earliest transducers used in measuring atmospheric pressure

is the barometer. A simple mercury barometer is constructed from inverting
a mercury-filled glass tube in a bath of mercury (Figure 4-1). When the tube
is tall enough, a vacuum is created on top of the mercury column. The
atmospheric pressure Palm is calculated from the height of the mercury col-
umn h by the equation:
Pat, = pgh
where p is the density of mercury (equal to 13.6 X 10' kg/m3),
g is the acceleration due to gravity (equal to 9.8 m/s'", and
h is the height of the'mercury column measured in meters.
The S.I. unit of pressure is Pascal (Pa). At standard atmospheric pressure
(STP), one atmosphere is equal to 101 kPa at temperature 273 K (0°C);the
mercury column height h calculated from equation (1) is 760 mm. Therefore,
at STP, one atmospheric pressure corresponds to a mercury column height
of 760 mm, or simply 760 mmHg. If one uses water instead of mercury in
the column, using 1,000 Kg/m3 as the density of water, the height of the
water column is calculated to be 10.3 m using the same equation. In addition
to the S.I. unit, the two most commonly used pressure units in biomedical
measurement are mmHg (e.g., for blood pressure measurement) and cmHzO
(e.g., for respiration pressure measurement). The conversion factors of some
frequently used units in pressure measurement are tabulated in Table 4-1.
 Pressure and Force Transducers

Figure 4-1. Mercury Barometer.

Table 4-1.
Pressure Unit Conversion
- - - - - -- - - -

-
&O
- -
psi -
milli Bar
- -

7.50 x 10-"-02 x 10.' 1.45 X 1 0'' 1.00 x 10"

mmHg
cmHsO
psi
milli Bar
- - -
100

Example 7

i) What is 760 mmHg in kPa?

ii) What is 2 cmHzO in psi?

Solution:

i) In the first column of Table 4-1, the multiplication factor to convert

mmHg to Pa is 133. Therefore, 760 mmHg = 133 x 760 Pa =
101,080 Pa = 101.08 kPa.
ii) In the fourth column of Table 4-1, the multiplication factor to con-
vert cmHzO to psi is 1.42 X lo-'. Therefore, 2 cmHzO = 2 X 1.42 X
lo-' = 2.84 X psi.

In physiological pressure measurement, as the human body is constant-

ly under one atmospheric pressure, the pressure measured is expressed as
the pressure above atmospheric pressure instead of the absolute pressure Pabs.
The pressure above atmospheric pressure is called gauge pressure Pgauge.The
 Biom~dr'dDeuice Rchology: Priadples and Daign

relationship between absolute, atmospheric, and gauge pressures is given by

the equation:

Whereas a barometer is used to measure the atmospheric pressure, a

manometer can be used to measure the pressure from any source. Figure 4-2
shows a manometer constructed from a liquid-Filed 13-shaped glass tube with
one end connected to a h w n clQniEtantpcmeSLUSGE A 4 tbe 6 % ertd
connected to the source to be measured. If the source pressure is Pa,using
the Pascal principle, the relationship between Pa,A+ the difference in lQdd
cahmn hefght 4 and the ffquid density p is given by:
Pa = Pb + pgh (3)
In measuring gas pressnre, one must remember that the pressure in a
closed cantainer changes with its temperature. The relationship between the
g a pressure P, its volume and the temperature Tis governed by the gas
law*which states that:

where PLR is a constant.

Figure 4-2. Manometer.

MECHANICAL PRESSURE GAUGES

Three types of mechanical pressure-sensing elements are often used in

gas pressure measurement: bourdon tube, diaphragm, and bellow. A bour-
don tube pressure transducer is shown in Figure 4-3; a bourdon tube is a hol-
 Pressure and Force Transducers

tube

Pressure being measured -

Figure 4-3. Bourdon Tube Pressure Gauge.

low metal coil with an oval-shaped cross section. When the pressure inside
the coil increases, the pressure creates a force to unwrap the coiled tube. A
mechanical linkage translates this movement into a pressure readout scale.
The scale is calibrated against known pressure sources.
Figure 4-4 shows a diaphragm pressure transducer. When the pressure
in the measurement chamber increases, the higher pressure on the measure-
ment side pushes the diaphragm outward. The movement of the diaphragm
is then converted to movement of a pointer needle to indicate the pressure
to be measured.

Diaphragm
/ Readout scale

Figure 4-4. Diaphragm Pressure Gauge.

In Figure 4-5, a bellow is used as the sensing element. Similar to the

diaphragm transducer, the bellow extends (or shortens) when the pressure
inside the bellow becomes higher (or lower). Through a mechanical linkage,
 Biomedical Deuice Technology: Principles and Design

the bellow extension creates a motion on the pointer to indicate the pressure
exerted on the bellow.

\ Readout scale

Bellow

Pressure being
measured
- >'
Figure 4-5. Bellow Gauge.

STRAIN GAUGES

Passive resistive sensors are commonly used in physiological measure-

ment. The resistance of a resistor wire is given by the equation:

where L is the length of the resistive wire, A is the cross-sectional area of the
wire, and p is the resistivity.
A passive transducer can use one of the three parameters in equation (4)
(length, area, and resistivity) to change its resistance, and thereby transduct
the physiological event. The two most important examples are the transduc-
tion of pressure and the transduction of temperature. For example, a device
known as a strain gauge can be used to measure force or pressure. A strain
gauge is either a length of a thin conductor or a piece of semiconductor that
is stressed or compressed in proportion to an applied force or pressure. The
extension or contraction of the strain gauge element results in a change of the
resistance.
Consider a piece of conductor wire, the resistance of which is given by
 !@,
Pressure and Force Transducers

equation (4). As the material is stretched, its length will increase and its cross-
sectional area will decrease. Both of these changes will cause an increase in
the resistance of the conductor. Alternatively, if the material is compressed,
the length will decrease and the area will increase and therefore will cause a
reduction in the conductor's resistance. In fact, for most materials, the unit
increase in length is proportional to the unit decrease in diameter. This pro-
portional constant is called Poisson's ratio v and is material-dependent.
Poisson's ratio is defined as:

v=-
ALI~
where is called the lateral strain, and AL/L is the axial strain.
Both the lateral and axial strains have no unit as the numerator and
denominator in equation (5) have the same unit. However, the axial strain is
often given a unit E , and due to its small value, it is often expressed in p ~ .
If one takes a partial derivative of R in equation (4), it becomes

Dividing both sides of the equation by R (the right-hand side by R and

the left-hand side by p -)L gives
A
AR - AL -
AA + Ap
R L A p

For a cylindrical resistant wire, A = where D is the diameter of the

4
wire. Taking the partial derivative of this equation and dividing both side by
A gives

Substituting equation (8)into equation (7) gives

Substituting equation (5)into equation (9) gives

Equation (10) shows that both the change in resistivity and the change in
length of the wire affect the resistance. The unit change of resistance per unit
 Biomedical Deuice Technology: Princ$les and Design

change in length of a strain gauge is called the gauge factor (G.F.).

AR/~
G.F. = -
AL/~
From equations (10) and (ll),the gauge factor G.F. is equal to

The G.F. value is different from one material to another. For a metal
strain gauge, - AP is zero and the G.F. is between 2 and 4. For a semicon-
P
ductor or piezoelectric strain gauge, -AP is nonzero. The G.F. of a piezo-
P
electric stain gauge can be several hundreds.

Example 2
For a metal wire resistive strain gauge with nominal resistance Ro =
120.0 R and G.F. = 2.045, find the change in resistance if an axial strain of
7,320 J . L ~ is applied to the strain gauge.

Solution:
From equation 11
AR/~
G.F. = -
ALI~
AR = R X G.F. X AL/L
AR = 120.0 x 2.045 x 7320 x 10-0= 1.796 n

A strain gauge may be either bonded or unbonded. A bonded strain

gauge means that the wire or semiconductor material is attached to a flexi-
ble bonding material, such as paper or plastic. It is then cemented to the sur-
face of the structure to measure the strain. Unbonded strain gauges use posts
to hold the ends of the gauge. The posts are attached to the structure to be
measured. Figure 4-6 shows bonded and unbonded metal wired strain
gauges.
Other than metal wire strain gauges, a common transducer element in
pressure measurement is the diaphragm piezoresistive strain gauge. It is fab-
ricated using semiconductor technology in which a layer of piezoresistive
material is deposited on a flexible diaphragm and lead wires are bonded to
 Pressure and Force Transducers

Fixed post Moveable post

n
w
n

/ a) Unbonded
Strain gauge wire \ 7Bonding material

b) Bonded
Figure 4-6. Wired Strain Gauges.

the strain gauge (Figure 4-7). The gauge resistance changes as the strain ele-
ment on the diaphragm is deformed by the applied pressure. As the gauge
factor of piezoresistive material is much higher than that of metal, it creates
a much more sensitive pressure transducer than metal wire stain gauges.
Furthermore, by controlling the position of the piezoresistive deposit on the
diaphragm and its shape, it can produce a linear output response to the
applied pressure.

- ,
Solid frame ,

wires

f -<I Diaphragm

Figure 4-7. Diaphragm Piezoresistive Strain Gauge.

While the focus has been on the use of the strain gauge for the transduc-
tion of pressure, it does have many other biomedical applications. An inter-
esting application in cardiology technology is the use of strain gauges to mea-
sure cardiac contractility. In this application, which is used for research pur-
poses, a bonded strain gauge is sutured directly to the ventricular wall of the
 Biomedical Device Technology: Principles and Design

heart. The contractile force of the muscle fibers causes a change in the strain
gauge resistance, which is then measured, processed, and displayed. Another
example is a load cell to measure the body weight of a patient undergoing
renal dialysis. A load cell is a transducer that converts a load (or force) act-
ing on it to an analog electrical signal. An example of a load cell is shown in
Figure 4-8, where the deformation of the bonded strain gauge on the beam
is converted to a change in resistance.

Force

Bonded strain gauge

I
I
#

, I

Anchor
Figure 4-8. A Load Cell.

PIEZOELECTRIC PRESSURE TRANSDUCERS

Some materials such as quartz crystal and certain ceramic materials gen-
erate opposite electric charges at their surfaces when subjected to mechani-
cal strain. Also, these materials produce physical deformation when an elec-
tric potential is applied. The charge Q produced is proportional to the
applied force F (which created the strain on the material). The proportional-
ity constant is called the piezoelectric constant K Consider the circular disc
piezoelectric transducer in Figure 4-9 with thickness d and cross-sectional
area A. The charge Qcreated on the surface of the transducer is:
Q=kXF
Assuming the transducer is a thin circular disc, with charge residing on
the surface, the disc can be considered as a parallel plate capacitor with
EA
capacitance C = -. The voltage across the capacitor
d
 , , 1111
Pressure and Force Transducers

It shows that the voltage measured across the surface of the transducer is
proportional to the applied pressure.

Applied force

Piezoelectric
element
Figure 4-9. Circular Disc Piezoelectric Element.

A biomedical application of such a transducer is the ultrasound trans-

mitter and receiver. In an ultrasound receiver, the sound pressure to be
detected produces some mechanical strain on the piezoelectric element,
which generates a potential difference whose amplitude varies according to
the sound pressure. In an ultrasound transmitter, the electric signal imposes
on the transducer produces deformation according to the amplitude and fre-
quency of the signal. This physical deformation creates mechanical pressure
waves in the medium in which the transducer element is submerged.

A pressure transducer with nominal resistance Ro = 100 R (at zero pres-

sure) and sensitivity S = 3.00 R per 100 mmHg is placed on one arm of a
Wheatstone bridge as shown in the figure below. If the resistors Rr, R2, and
fi are all 100 R and the excitation voltage VE = 5.00 V, find the change in
output voltage Iro when the pressure is increased from 0 to 60.0 mmHg.

Solution:
When P = 0 mmHg, Rt = RI = R2 = R3 = 100 IR, the bridge output Vo
= 0.0 V (balanced bridge).
When the applied pressure P = 60.0 mmHg, the change in resistance of
the transducer A& = S X A P = 3.00 IR/100 mmHg X 60.0 mmHg = 1.80 IR.
 Biomedical Device Technology: Principles and Design

Therefore, Rr = 100 + 1.80 = 101.8 0, substituting into the bridge circui

gives
100
Vo = [& - &]= 5.00[ 101.8 + I00
-
loo
100 + LOO]
= -0.0223V = -22.31111
 Chapter 5

TEMPERATURE TRANSDUCERS

OBJECTIVES

Explain the IPTS temperature scale.

Perform temperature unit conversion.
State common temperature measurement terminology
Describe the principles of nonelectrical temperature gauges.
Analyze the characteristics of RTDs, thermistors, thermocouples, and
I C temperature sensors.
Define errors due to lead resistance and self-heating effect in tempera-
ture measurement applications.
Explain methods to linearize thermistor characteristics.

CHAPTER CONTENTS

1. Introduction
2. Reference Temperature and Temperature Scale
3. Nonelectrical Temperature Transducers
4. Electrical Temperature Transducers
5. Resistance Temperature Devices (RTDs)
6. Lead Errors
7. Self-Heating Errors
8. Thermistors
9. Thermocouples
10. Integrated Circuit Temperature Sensors
11. Comparison of Temperature Sensors' Characteristics
 Biomedical Device Technology: Princij6les and Design

INTRODUCTION

Temperature transducers have widespread applications in biomedical

instrumentation. They can be used in direct body temperature measurement,
in patient assessment, or to ensure patient safety by monitoring performance
of machine parameters such as the heating fluid temperature of a blood
warmer. Temperature transducers that are designed to provide temperature
readings are called thermometers. This chapter covers concepts in tempera-
ture measurement and the principles of some commonly used temperature
transducers.

REFERENCE TEMPERATURE AND TEMPERATURE SCALE

Unlike most other physical quantities, one cannot add temperatures

together as one would add lengths to measure distances. We must rely on
physical phenomenan to establish observable and consistent temperature
references. For example, the freezing point and boiling point of water at one
atmospheric pressure are assigned as 0' and 100°, respectively, in the Celsius
(or centigrade) scale and as 3 2 O and 212' in the Fahrenheit scale. The entire
temperature scale can be constructed by making interpolation of these fixed
temperature references. Table 5-1 shows the reference temperatures on
which the International Practical Temperature Scale (IFTS) is based.
Interpolation of these references is performed by temperature transducers.
Common units of temperature measurement are degree Celsius ( O C ) in
the Celsius scale, degree Fahrenheit ( O F ) in the Fahrenheit scale, and the

Table 5-1.
IPTS-68 Reference Temperatures
- - -

Reference Point - -
-
K
Triple Point of Hydrogen
Liquid/Vapor Phase of Hydrogen at 25/76 Std.
Boiling Point of Hydrogen
Boiling Point of Neon
Triple Point of Oxygen
Boiling Point of Oxygen
Triple Point of Water
Boiling Point of Water
Freezing Point of Zinc
Freezing Point of Silver
Freezing Point of Gold
-
 Temperature Transducers

Kelvin (K) in the absolute scale. For a temperature of A°F, the equivalent
temperature reading in the Celsius scale is P C , where

9B
or, in reverse, A = -
5
+ 32.
The same temperature in Kelvin (CK) is given as
C =B + 273.15.
Example 7
The body temperature of a patient is measured to be 37.0°C, what are the
temperature readings in the Fahrenheit and absolute scales.

Solution
9 B
From equation (I), A = - 9 X 37.0 + 32 = 98.6OF.
+ 32 = -
5 5
From equation (2), C = B + 273.15 = 37.0 + 273.15 = 310.2 K.

NONELECTRICAL TEMPERATURE TRANSDUCERS

Fluid-in-glass thermometers have long been used to measure tempera-

ture. It has been the most popular type of thermometer in body temperature
measurement until recently. Mercury-in-glass and petrolate-in-glass are the
two most common types in this category. A fluid-in-glass thermometer is
shown in Figure 5-1. It consists of a glass tube (or stem) with a uniform
lumen connected to a reservoir. The top of the tube is sealed after air is
removed. When the temperature increases, the fluid in the reservoir expands
and pushes up the fluid level in the stem. The position of the fluid meniscus
indicates the temperature under measurement. Two temperature references
(e.g., water freezing and boiling points) are chosen to calibrate two points on
the stem of the thermometer. Mercury-filled thermometers usually have a
temperature measurement range from -40 to +900°C, while red-dyed petro-
late can measure from -200 to +260°C. In medical applications, some glass
thermometers are TeflonTM encapsulated to avoid shattering.
Bimetallic sensors are often used in temperature gauges as thermal
 Biomedical Device Technology:Princt$les and Design

Graticule

Figure 5-1. Fluid-in-Glass Thermometer.

switches. Two metals with different temperature expansion coefficients are

bonded together to form a bimetallic strip. When the temperature changes,
the differential expansion causes the strip to bend. The degree of bending
can be calibrated and used in temperature measurement. Figure 5-2 shows
a simple bimetallic strip, and Figure 5-3 shows a dial-type temperature gauge
with the bimetallic strip formed into a helical coil to produce rotational
motion with changes in temperature. Depending on the type of metals used,
bimetallic gauges can measure temperature in the range of -40 to 1500°C.

Metal B
Cool bimetallic strip Heated bimetallic strip
Figure 5-2. Bimetallic Temperature Sensor.

Another nonelectrical temperature sensor is constructed by sandwiching

a liquid crystal material between an adhesive backing and a transparent
MylarTM film. The liquid crystal can be made to reflect different colors of light
at different temperatures. Liquid crystal thermometers provide fast (1 to 2
seconds respond time) visual indication of temperature with a range from 0
to +60°C. Strip thermometers are available to measure body temperature.
 Temperature Transducers

Stem Shaft

Figure 5-3. Dial-Type Temperature Gauge.

ELECTRICAL TEMPERATURE TRANSDUCERS

In general, electrical transducers can be divided into two types-passive

and active. Resistance temperature devices and thermistors, both of which
are passive transducers, are covered in this chapter. Thermocouples and inte-
grated circuit sensors are covered as examples of active temperature trans-
ducers.
In addition to other instrumentation specification parameters covered in
Chapter 2, some parameters that are commonly referred to in temperature
measurement applications are:
Operating range-the temperature between two limits within which the
characteristics conform to the specifications.
Stability-the quality of the sensor to maintain a consistent output
when a constant input is applied. The shelf life of a disposable trans-
ducer is an indication of the stability of the sensor.
Dissipation constant-the power in milliwatts required to raise the sen-
sor by 1°C above the surrounding temperature. For example, a ther-
mistor may be listed to have a dissipation constant of 1.0 mW/OC in
still air.
Respond time-the time required to reach a certain percentage of the
steady state output followed by a step input. The time to reach 63.2%
is referred to as one time constant. For most temperature transducers,
approximately five time constants are necessary to reach 99010 of the
steady state output.
 Biomedical Device Technology: Principles and Design

RESISTANCE TEMPERATURE DEVICES (RTD)

For a metal conductor, the resistance increases with temperature.

Therefore, one can measure the resistance of a metal wire to determine its
temperature. Figure 5-4 shows the normalized resistance-temperature char-
acteristics of metal RTDs, including nickel, copper, platinum, and tungsten.
RT is the resistance of the metal wire in (n at temperature T°C, and Ro is the
resistance of the metal wire at To°C. By selecting different metals, RTD can
measure temperature from -200 to over +800°C. As shown in Figure 5-4,
despite its low sensitivity, the platinum RTD shows an almost linear resis-
tance-temperature relationship and therefore can be approximated by:
RT = Ro (1 + aAT) (3)
where AT= T- Tq
To is usually specified at O°C,
and a is the fractional change in resistance per unit temperature.
The temperature coefficient is defined as

which is equal to aRo from equation (3).

I I
400 800
Figure 5-4. Characteristics of RTD.
 Temperature Transducers

Example 2
A platinum RTD is used in a bridge circuit as shown below. What is the
bridge output KI at 10°C, given the excitation voltage fi = 2.0 V, bridge
resistance R = Ro = 100fl at O°C, and a = 0.00385/OC.

Solution
Let AR = RT - Ro,
at 10°C, AR = Ro ( 7 + aA7) - Ro = a U T = 0.00852*700*70 = 3.850
for the Wheatstone bridge circuit.

Substituting R = 70032, AR = 3.850, and VE = 2.0 Vinto the above gives

Vo = -78.9 mV

Example 3
A piece of platinum RTD wire has a resistance of 100.0fl at O°C.
Find (a) its temperature coefficient and (b) its resistance at 180.0°Cgiven that
a = 0.00385/OC for platinum RTD.

Solution
(a) Temperature coefficient = aRo = 0.00385 X 100 = 0.385fl/OC
(b) Using equation (3), Rixr, = 100 (1 + 0.00385*189) = 171.8 f l
The value of a quoted in this example is according to the DIN Standard
(European). In the United States, a is equal to 0.00392/OC for a standard
platinum RTD. Usually, the RTD characteristics are specified in a lookup
table of resistance against temperature supplied by the manufacturer.

LEAD ERRORS

It should be noted that RTDs in general have law sensitivities. That is,
the output resistance changes onIy siightly for a relatively large change in
temperature. For the platinum RTD in the above example, the temperature
sensitivity (ortemperature coefficient) b e q d t~ O.385VI0C.In tenaperare
measurement using RTD, one must be aware of the errors introduced to the
mewurement from inadvertent additions and changes in the overall mea-
sured resistance.

Example 4
In a temperature measurement experiment using a platinum RTD, a pair of
long lead wires are used to connect the RTD to an ohmmeter. If ol =
0.00385/'C and each lead wire has an overall resistance of 0.5Q find the
lead error.

Solution
+
The lead wires add a total resistance of 0.5 0.5 = 1.OSZ to the RTD.
The temperature error TEdue to the lead resistance (or simply called lead
error) is
10 = 2.6%
TE= -
0.385
which is a very significant error in medical applications.

Example 5
For the RTD application in Example 2, if each lead wire to the RTD has a
resistance RL of 0.50Q what is the bridge output?

Solution
With the lead resistance RL, the effective resistance RE across the bridge arm
+
becomes RT ZRr, substituting into
 Temperature Transducers

Substituting R = 7000, AR = 3.850, and VE = 2.0 Vinto the equation

vo = - 2 X 3.85 X 2 X 0.50 X - 2.0 gives

100 + 3.85 + 2 X 0.50 2
Vo = -23.7 mV.
This result shows a -25% error compares to the bridge output (- 18.9 mV)
without lead resistance in Example 3.

In any resistive transducer applications, lead resistance can introduce sig-

nificant errors in the measurement. Lead errors can be accounted for if one
knows the exact resistance of the lead wires. However, this may not be pos-
sible in some applications. Below are some methods used to minimize lead
errors.

Three-Wire Bridge
One method to reduce the lead error is to use a 3-wire bridge circuit.
Instead of using a conventional 2-wire RTD, a 3-wire RTD is used as shown
in Figure 5-5.

F
' 3-Wire RTD
\.----

Figure 5-5. 3-Wire Wheatstone Bridge Circuit.

For the circuit in Figure 5-5, if we are using a high input impedance volt-
meter to measure Vo, there will be no voltage drop across the middle lead
(RL) since the middle lead does not carry any current. Given RT = R AR, +
the bridge output becomes
 Biomedical Device Technology:Principles and Design

For the 3-wire RTD application in Figure 5-5, using the same circuit values
as Example 5, that is, R = 100fi, RL = 0.05, AR = 3.85fi (at 10°C), and VE
= 2.0 V, what is the bridge output?

Solution
Substituting R = 100 IR, AR = 3.85 IR, RL = 3.85 IR, and VE = 2.0 V into
the equation.

vo = - 2R + ~ARR +L AR X k,
which gives
2

This result shows a mere 0.5010 error compares to bridge output (- 18.9 mV)
without lead resistance (Example 2). This is a significant improvement to the
2-wire RTD bridge circuit in Example 5 (-25% error)!

To totally eliminate the error due to lead resistance, a 4-wire RTD with a
high impedance voltmeter and constant current source connected as shown
in Figure 5-6 is needed. The RTD resistance is simply calculated by divid-
ing the voltmeter reading to the current from the constant current source.

RTD

Figure 5-6. 4-Wire RTD Circuit.

 Temperature Transducers

SELF-HEATING ERRORS

An RTD is a resistor whose resistance changes with temperature. As we

all know, when a current Ipasses through a resistor R, it will produce heat at
a rate equals to I Z R This heat will raise the temperature of the RTD and
therefore will create an error in the measured temperature. Dissipation con-
stants are specified for all resistive-type temperature transducers. The dissi-
pation constant specifies the temperature error caused by the heat generated
in the transducer due to externally applied current. The value of dissipation
constant depends on the type of transducer element, its packaging, as well as
the medium in which it is being used.

Example 7
A 100a platinum RTD in a bridge circuit is used to measure air temperature.
The bridge excitation is 5.0 V and all the bridge resistors are IOOR. Calculate
the error due to self-heating if the dissipation constant of the RTD in still air
is 5.0 mW/OC (or 0.20°C/mW).

Solution
When the bridge is balanced, i.e. RT = R = 100a.
The voltage VTacross RT is 0.5 times the excitation voltage VE.The power
dissipated Pr in the RTD is therefore given by

Therefore, self-heating error = 62.5 mW = 12.50C,

5.0 mW/OC
 Biomedical Device Technology: Principles and Design

which means that the RTD will read 12.5OC higher than the correct temper-
ature due to the self-heating effect.
From the above example, one can see that in order to reduce the self-heat-
ing error, one should reduce the excitation voltage or increase the bridge
resistance.

THERMISTORS

Another common passive resistive temperature transducer is the ther-

mistor. Thermistors are made from semiconductor materials such as oxide of
nickel, manganese, iron, cobalt, or copper. As a thermistor is a semiconduc-
tor, it has nonlinear resistance temperature characteristics. Thermistors can
have a positive or negative temperature coefficient and an operating range
from -80 to + 150°C. Thermistors have higher nominal resistance than
RTD (ranges from 1 kR to 1 M a ) as well as higher sensitivity (from 3 to
5%/OC). Although not as good as RTD, it has very good long-term stability
(less than 0.2% drift in resistance per year). The most common type of ther-
mistor used in medical instrument is the YSI 400 series thermistor (YSI
stands for Yellow Spring Instrument, which is a manufacturer of temperature
transducers), which has a negative temperature coefficient (Figure 5-7). The
characteristics of a thermistor are available in manufacturer's lookup tables.
Also, most negative temperature coefficient thermistors can be approximat-
ed by:

where Rris the thermistor resistance in R at temperature T,

Ro is the thermistor resistance in R at temperature To,
Tand To are absolute temperature in K (note that Tis in OC in the RTD equa-
tion), and a is the thermistor material constant in the range of 2,500 to 5,000
K.
The temperature coefficient
(d&-
dT)
can be obtained by differentiating
equation (4) with respect to temperature T, which gives:
d[e8(k-k)]
(%) Ru =
dT
= Roe
 Temperature Transducers

T(K)
Figure 5-7. Thermistor Characteristics.

For a thermistor with p = 4,000 K, find (a) the thermistor resistance, and (b)
the temperature coefficient at 37.0°C given the Ro = 7355fl at O°C.

Solution
(a) Using equation (4)

(b) Temperature coefficient = - RTB - -1,280 X 4,000 = -53.3

T" (37 273)" +
One of the major drawbacks for thermistor is its highly nonlinear char-
acteristics. Much research was done to produce thermistors with linear resis-
tance temperature characteristics. A rather efficient method to produce a
piecewise linear characteristic within a narrow temperature region is
achieved by connecting a parallel resistor RP to the thermistor. This combi-
nation produces an approximate linear region centered around the temper-
ature of interest Tm. The parallel resistor value RP is derived from finding
- the
( dT'
point of inflexion ==O of the resistor combination curve. Without going
into the detailed derivation, RP can be calculated from:
 Biomedical Device Technology: Principles and Design

A Thermistor
characteristic

, inflexion of
, linearized
Approx.
linear
region
.'. .

Figure 5-8. Thermistor Linearization.

The linearized characteristic using this method is shown in Figure 5-8.

Note that this method reduces the overall sensitivity of the transducer.

Example 9
A YSI 400 thermistor is used for body temperature measurement.
Choose a parallel resistor to the thermistor to linearize its resistance temper-
ature characteristic around the temperature of interest.

Solution
The normal body temperature is about 37OC (= 310 K). From the man-
ufacturer's lookup table of YSI 400 thermistor, R 3 7 = 1,355 R and P = 4,000
K. Substituting into equation (5)gives

Another method to obtain a linear relationship is by using two thermis-

tors, Figure 5-9a shows a YSI 700 series thermistor combination, and Figure
5-9b shows one example of using this thermistor to produce a linear resis-
tance temperature curve (Figure 5- 10).
Note that there are three lead wires for a YSI 700 Series thermistor (one
more than a 400 Series). There are other methods to linearize thermistor
characteristics. However, newer microprocessor-based devices with large
 Temperature Transducers

Thermistor pair

Figure 5-9. YSI 700 Series Thermistor and Application.

Figure 5-10. Linearized Characteristic of YSI 400 Series Thermistor Circuit.

rnemoq to s h e the entire thermistor lookup table have made linearfiation

less Important. Different sbes and shapes of thermistor probes are designed
for different applications.
THERMOCOUPLES

A thermocouple is an active temperature transducer that produces a

small unique voltage according to the temperature. A thermocouple consists
of two dissimilar metals joined together. Thomas Seebeck discovered this
property in 1821.
 Biomedical Deuice Technology: Princ$les and Desigrz

Seebeck Effect
When two wires of dissimilar metals are joined at both ends and one end
is at a higher temperature than the other, there is a continuous flow of cur-
rent in the wire (Figure 5-lla). This current is called the thermoelectric cur-
rent. When one junction is disconnected, a voltage called the Seebeck volt-
age can be measured across the two metal wires (Figure 5-llb).

Metal B (a)

Figure 5-1 1. Seebeck Effect.

For a small variation of the temperature difference between the hot and
cold junctions, the Seebeck voltage EAB is proportional to the temperature
variation AT with a proportional constant a called the Seebeck coefficient.
This relationship is represented by
EAB = CXAT
The Seebeck coefficient is different for different thermocouples. It is
small and varies with temperature. In addition, it has a nonlinear tempera-
ture characteristic. For example, the Seebeck coefficient for E-type (Chrome1
and Constanan) thermocouple varies from about 25 pV/OC at -200°C to 62
pV/OC at +8OO0C.
The thermoelectric sensitivity of a thermocouple material is usually
given relative to a standard platinum reference. Since the Seebeck coefficient
is temperature-dependent, the sensitivity is also temperature-dependent.
Table 5-2 shows the sensitivities of some thermocouple metals referenced to
platinum at O°C. The sensitivity at O°C of any combination of the materials
 Temperature Eansducers

can be obtained by taking the difference of the sensitivities of the materials

forming the thermocouple. For example, the sensitivity of chrome1 and con-
stanan (Type E thermocouple) is -25 - (-35) = 60 pV/OC. In temperature
measurement, one of the junctions is kept at a known (or reference) temper-
ature and the other junction is at the temperature to be measured. The ref-
erence temperature is often referred to as the cold temperature since ther-
mocouples are usually used for high temperature measurements.

Table 5-2.
Thermoelectric Sensitivity of Thermocouple Materials
- - - - - - - - -

Copper -
Chrome1 -
Constanan
-
Iron Platinum
Composition Cu 90 Ni/lO Cr 60 Cu/4O Ni Fe Pt
Sensitivity, pV/OC
- -
+6.5 +25
-
-35
-
+ 18.5
-
0

Thermocouples can be used to measure temperature from -200 to +2000

OC.In addition to its wide temperature range, it is rugged, accurate and can
be made to be extremely small in size. The output characteristics of some
common thermocouples are listed in Table 5-3 and shown in Figure 5-12.

Table 5-3.
Characteristics of Thermocouples
- -- - - -- - -

m e Name Temp Range

- - Output: mV
- -

T Copper/constanan -200 to +300 4.25

J Iron/constanan -200 to +1,100 5.28
E Chromel/constanan 0 to +1,100 6.30
K Chrornel/alurnel -200 to 1,200 + 4.10
S -
Platinum/pt rhodium - -
0 to +1,450 0.64
* Output is measured at 100°C with reference at O°C.

Empirical Laws of Thermocouples

The following three empirical laws are important facts that enable ther-
mocouples to be used as practical temperature measurement devices:
1. The law of homogeneous circuit
2. The law of intermediate metals
3. The law of intermediate temperature
The law of homogeneous circuit states that for a thermocouple of metals
A and B, the output voltage is not affected by the temperature along metal A
 Biomedical Device Technology: Principles and Design

500 1,cm 1,500 2,000

Figure 5-12. Thermocouple Output Characteristics.

or B. It is determined by the temperatures at the two junctions (Figure 5-13).

This is required as the metals or lead wires between the thermocouple junc-
tions will be at different temperatures than either of the junctions.

Metal A
:----:
: T3 :

Tl T2

Metal B V~~ = Vcd

Figure 5-13. Law of Homogeneous Circuit.

The law of intermediate metals states that the output voltage is not affect-
ed by the use of a third metal in the circuit provided that the new junctions
are at the same temperature (Figure 5-14). This allows lead wires (usually
copper wires) of a voltmeter to be connected to the thermocouple circuit to
measure its output voltage.
 Temperature Transducers

Metal C
Metal A

Metal B VCD = Vcd

Figure 5-14. Law of Intermediate Metal.

The law of intermediate temperature states that for the same thermocou-
ple, the output voltage V31 where the temperature at the junctions are T3 and
TI, respectively, is equal to the sum of the output voltages of two separate
measurements with junction temperatures T3, Tz, and Tz, TI (i.e., f i 1 = f i z +
EI)(Figure 5-15). This allows using a standard thermocouple lookup table
to calculate the junction temperature from the voltage measured.

Metal A Metal A Metal A

\

Metal B Metal B Metal B

Vxy = Vab + Vcd

Figure 5-15. Law of Intermedate Temperature.

In practice, the setup for temperature measurement using a thermocou-

ple is as shown in Figure 5-16a. The empirical law of intermediate metal
allows a third metal to connect the thermocouple to the voltmeter. It is
important that the connections between the lead wires and the thermocou-
ple metal wires be kept at the same temperature to avoid generating anoth-
er voltage to upset the measurement. These connections are bonded to an
isothermal block that has high thermal conductivity to ensure equal temper-
ature at the connection points.
 Biomedical Device Technology: Princ$les and Design

At the hot junction (temperature to be measured T m ) , the junction volt-

age referenced to O°C is equal to Vmo. At the isothermal block (at tempera-
ture Tr), the junction voltage referenced to O°C is equal to Vti). Using the
empirical law of intermediate temperature, we can write:
Vmr + KO= Vmo, where Vmr = Vis the thermocouple output voltage, or
V = Vmo - Vro (6)
If we know the reference temperature T7 and the output voltage, with the
thermocouple lookup table, we can use equation (6) to find the hot junction
temperature Tm.

Example 70
A J-type thermocouple is used to measure the temperature of an oven. If the
temperature of the isothermal block is at 20.0°C and the output voltage mea-
sured is 9.210 mV, what is the temperature of the oven?

Solution
Step 1
Using a J-type thermocouple lookup table, find the junction voltage KOat Tr
= 20°C. At 20.0°C, the junction voltage of a J-type thermocouple referenced
to O°C from the lookup table is 1.019 mV.
Step 2
Using equation (6), Vmo = V + KO= 9.210 + 1.019 = 10.229 mV.
The junction voltage at the hot junction referenced to O°C is 10.229 mV.
Step 3
From the thermocouple lookup table, at Vmo = 10.229, Tm = 190°C.
Therefore, the temperature of the oven is 190°C.

Metal A

block
Compensation
circuit
la) " lb)
Figure 5-16. Thermocouple Measurement Setup.
 Temperature Transducers

The setup in Figure 5-16b includes a temperature sensor (thermistor) to

measure the reference temperature. The compensation circuit produces a
voltage according to the thermistor resistance. The output voltage becomes
the algebraic sum of the compensation voltage and the junction voltage,
which can be converted to the junction temperature.
Thermocouples are available in different package formats for different
applications. The junction can be exposed, ungrounded, or grounded. An
exposed junction is recommended for the measurement of static or flowing
noncorrosive fluid where fast response time is required. An ungrounded
junction is recommended for measurements in corrosive environments
where it is desirable to have the thermocouple shielded from the corrosive
environment by the protective sheath and electrically isolated and shielded.
A grounded junction is recommended for the measurement of static or flow-
ing corrosive fluid. The junction of a grounded thermocouple is welded to
the protective sheath. As the magnitude of the thermocouple output voltage
is usually very low, care must be taken to avoid errors due to electrical inter-
ference and voltaic effects.

INTEGRATED CIRCUIT TEMPERATURE SENSORS

The last temperature sensor to be discussed in this chapter is the I C tem-

perature sensor. IC temperature sensors employ the temperature-dependent
properties of the PN junction to measure temperature. I C temperature sen-
sors are in general very accurate and linear. However, due to the fact that
they are constructed from semiconductor materials, I C temperature sensors
are not suitable for extreme temperature conditions. The operating range of
IC temperature sensors is from -55 to + 150°C. I C temperature sensors can
be current type or voltage type. In a current type sensor, the sensor will pro-
duce a current proportional to the temperature. In a voltage type sensor, the
output voltage is proportional to the temperature. An example of an I C tem-
perature sensor is the National Semiconductor LM50, which has a tempera-
ture sensitivity of 10.0 mV/OC, +0.8O/o maximum nonlinearity and only
requires a 4 to 10 V single power supply.

COMPARISON OF TEMPERATURE SENSORS'

CHARACTERISTICS

Among the temperature sensors discussed, thermistors are widely used in

patient temperature measurement due to its small size and high sensitivity.
 Biomedical Device Technology: Principles and Design

However, it may not be suitable for other applications. Table 5-4 compares
the characteristics of the four types of sensors discussed.

Table 5-4.
-
Comparison of Temperature Sensors Characteristics -

Thermbtor Platinum RTD

- -
Thermocouple IC Sensor
Sensitivity 3 to 5 KW°C 0.00385/°C 7 to 60 pV/OC 1 pAl0c
Linearity Poor Good Fair Very good
Stability Stable Very stable Stable Less stable
Power Required Yes Yes Self powered Yes
Min. Practical Span 1OC 25OC 100°C 25OC
Temp. Range, OC - 100 to +250 -200 to +750 - 100 to +2,000 -55 to + 150
Reference Required No No Cold junction No
Ruggedness Very rugged Rugged Very rugged Rugged
Repeatability Very good Very good Fair Good
Hysteresis Low Low High Low
- -- - - - -
 Chapter 6

MOTION TRANSDUCERS

State the relationships between displacement, velocity, and accelera-

tion.
Derive the principles of resistive, capacitive, and inductive displace-
ment transducers.
Describe application examples of resistive, capacitive, and inductive
sensors in linear and angular displacement measurements.
Explain the principles of Hall effect sensors.
Describe application examples of Hall effect sensors in linear and angu-
lar displacement measurements.

CHAPTER CONTENTS

1. Introduction
2. Resistive Displacement Transducers
3. Inductive Displacement Transducers
4. Capacitive Displacement Transducers
5. Hall Effect Sensors

The measurements of displacement, velocity, and acceleration are often

performed in biomedical applications. Measuring the motion of the heart
 Biomedical Device Technology: Princ$les and Design

wall during open heart surgery, quantifying the frequency and magnitude of
hand tremor for a patient suffering from Parkinson disease, and monitoring
the position of a surgical retractor are a few examples of the applications.
The displacement x (in m), velocity v (in m/s), and acceleration a (m/s') of an
object are related by the following equations, where t is the time in seconds:

- - d'x
a = - -dv
dt dt'
Therefore, if the displacement versus time of an object is recorded, its
velocity and acceleration can be calculated. In a similar manner, if the angle
versus time is recorded, the angular velocity and angular acceleration can be
computed. This chapter focuses on the measurement of displacement includ-
ing linear (or translational) and angular displacement.

RESISTIVE DISPLACEMENT TRANSDUCERS

A passive displacement transducer is a device such that a displacement

causes a related change in a circuit parameter such as resistance, inductance,
and capacitance.
For a resistive wire, the resistance R is given by

where p = the resistivity of the wire material,

L = the length of the wire, and
A = the cross-sectional area of the wire.
An example of a linear displacement transducer using the above proper-
ty is shown in Figure 6-la, where the displacement to be measured is linked
to the wiper of the potentiometer. In this example, the measured resistance
is directly proportional to the displacement. A practical displacement trans-
ducer is shown in Figure 6-lb. Figure 6-2 shows an example of a resistive
angular motion transducer.
 Motion Transducers

Resistive wire
Mechanical linkage
to measurement site

Displacement x

*
Displacement x
T'
Resistive element housing
\
Lead wires

Figure 6-1. Linear Displacement Transducer.

Mechanical linkage
to measurement site

Figure 6-2. Resistive Angular Motion Transducer.

INDUCTIVE DISPLACEMENT TRANSDUCERS

The inductance L of a solenoid is given by

where N = the number of turns of the solenoid,

p = the effective permeability of the material inside the solenoid core,
A = the cross-sectional area, and
I = the length of the solenoid.
 Biomedical Device Technology: Princ$les and Design

A displacement transducer can be made by linking the change in dis-

placement with one of the parameters in equation (2). A simple example is
shown in Figure 6-3a, where the ferromagnetic core of the solenoid is con-
nected to the object of which the motion is to be measured. The effective per-
meability of the solenoid is changed by the displacement of the core in the
solenoid. The direction and magnitude of the displacement can thus be
found by measuring the inductance of the solenoid. Another example is
shown in Figure 6-3b, where the displacement is measured by measuring the
changes in magnetic flux caused by the motion of the magnetic material.

Mechanical link to
measurement site

00000000
I
J
d *
f Displacement x

Solenoid coil Magnetic core

Displacement x

f---+

Magnetic
material

Figure 6-3. Inductance Displacement Transducer.

Figure 6-4a shows a displacement transducer called the linear variable

differential transformer (LVDT).It consists of a primary winding L1 and two
secondary equal but oppositely wound windings L2 and L3. When Ll is con-
nected to an AC excitation and the core is right at the middle position, the
induced voltages of the two secondary windings are of equal magnitude and
hence cancel each other. The output of the LVDT is therefore zero. When
 Motion Transducers

the core is off-centered, the induced voltage of winding L1 is different from

that of L2. A nonzero output voltage Vo appears at the output. f i is propor-
tional to the distance of the core away from the center position. The input-
output characteristic is as shown in Figure 6-4b. In addition to its linear char-
acteristic, an LVDT theoretically has infinitely small resolution. Since f i is a
sinusoidal voltage, the output is always positive. A phase detector is therefore
required to determine on which side of the center the magnetic core is locat-
ed,

Mechanical link to

Core position x
Core

Figure 6-4. Linear Variable Differential Transformer.

CAPACITIVE DISPLACEMENT TRANSDUCERS

For a parallel plate capacitor, the capacitance Cis given by:

where E = the permittivity of the dielectric,

A = the overlapping area of the plates, and
d = the distance between the two plates.
A capacitive displacement transducer element converts a change in
either the permittivity E, the area A, or the plate separation d into a change
in the value of the capacitance C. The change in capacitance will in turn
cause a change in an electrical signal. For example, one of the plates can be
moved together with the object to be measured to provide a signal accord-
ing to the displacement of the object (Figure 6-5a). In Figure 6-5b, the per-
 Biomedical Device Technology: Principles and Design

mittivity E is changed by the movement of the dielectric linked to the dis-

placement to be measured. Figure 6-5c shows how the change in the over-
lapping area A can be tied to the displacement. A modification of that can be
used for angular displacement measurement (Figure 6-5d).

Moving plate Capacitor plates

Moving plate
mechan~callyl~nked
to measurement slte

Fixed plate Displacement x

Dielectric

Fixed plate

\I Fixed
plate

Moving plate Displacement x

(4 td)
Figure 6-5. Capacitive Displacement Transducers.

HALL EFFECT SENSORS

Hall effect sensors are found in many medical devices. For example, they
are used as a proximity sensor in the door switch of an infusion pump. Figure
6-6 illustrates the setup of a Hall effect sensor. The Hall element is a rectan-
gular plate of metal or semiconductor material such as bismuth or tellurium.
When the current carrying plate is placed in a magnetic field perpendicular
to the plate, a voltage (called the Hall voltage) will be induced at the side
faces of the plate in a direction perpendicular to the current in the plate. The
Hall voltage Eh is given by the equation:
KIB
Eh = -
t
 Motion Transducers

where I = the current flowing through the Hall plate,

B = the magnetic field perpendicular to the plate,
t = the thickness of the plate, and
K = the Hall coefficient, which is dependent on the material.

Figure 6-6. Hall Effect Transducer.

In practice, the current I is often held constant. From equation (4), En is

proportional to the magnetic field strength B perpendicular to the plate. Hall
effect sensors are commonly used in magnetic field measurements with sen-
sitivity of a few mV per kGauss. A Hall effect sensor can easily be adapted
to become a position or displacement sensor. Figure 6-7 shows several
setups using Hall sensors as displacement transducers.
In Figure 6-7a, the Hall sensor is positioned between the two north poles
of the magnets and supplied by a constant current. If the magnets are iden-
tical, the magnetic field strength is zero right at the middle and therefore the
output of the sensor (Hall voltage) is zero. The magnetic field strength at both
sides of the center varies with the distance from the center. By mechanically
connecting the object to be measured to the sensor, the motion of the object
can be monitored by measuring the Hall voltage. In the setup as shown in
Figure 6-7b, the Hall effect sensor is placed stationary between the poles of
a magnet. The magnetic flux in the magnet and therefore the magnetic field
at the Hall sensor varies with the air gap between the magnet and the ferro-
magnetic plate. Motion of the object can thus be measured by the Hall sen-
 Biomedical Device Technology: Princ$les and Design

Displacement x

Permanent
magnet
*
D~splacementx
(b)
displacement

(c)
Figure 6-7. Applications of Hall Effect Devices.

sor by connecting the ferromagnetic plate to the object. In Figure 6-7c, the
magnetic field between the poles is constant. However, when the Hall effect
sensor rotates, the magnetic field perpendicular to the surface of the sensor
varies by a factor equal to sin 0, where 0 is the angle between the magnetic
axis and the surface of the sensor. In essence, a displacement sensor can be
constructed by linking the object to a setup such that the Hall effect sensor
element is exposed to changing magnetic field strength caused by the motion
of the object.
 Chapter 7

FLOW TRANSDUCERS

Explain the differences between laminar and turbulent flow.

State Bernoulli's equation and Poiseuille's law.
Define viscosity and Reynolds number.
Derive the Venturi tube equation and explain the principle of operation
of Venturi tube, orifice, pitot tube, and rotameter flow meters.
Explain the operation of turbine and paddle wheel flowmeters, hot wire
anemometers, electromagnetic and ultrasound flow sensors.

CHAPTER CONTENTS

1. Introduction
2. Laminar and Turbulent Flow
3. Bernoulli's Equation
4. Flow Transducers

The cardiovascular system consists of the heart, which generates pressure

to circulate blood through a network of blood vessels around the body. The
blood pressure and flow velocity are different at different locations of the cir-
culatory system. Chapter 4 covers pressure transducers that can be used to
measure blood pressure, while this chapter covers flow transducers to mea-
 Biomedical Device Technology: Principles and Design

sure blood flow velocity. Physiological parameters in the flow measurement

category include, blood flow, respiratory gas flow, and urine flow.
Flow measurement includes measuring the velocity v (m/s), volume flow
rate ()(US), or mass flow rate F (kg/s). The mass flow rate is defined as the
mass m of the fluid that flows through a cross section of the vessel per unit
time t. Since mass is equal to volume Vtimes density p, the mass flow rate is
given by
F=-=-- p X V - p X A x l = p x A x v = p x Q (I)
t t t
where A = the cross section of the vessel, and
1 = the distance of the fluid path.

LAMINAR AND TURBULENT FLOW

There are two types of fluid flow: laminar flow and turbulent flow.
Laminar (or streamline) flow is smooth flow such that neighboring layers of
the fluid slide by each other smoothly. It is characterized by the fact that each
particle of the fluid follows a smooth path and the paths do not cross over
one another; such a path is called a streamline. Laminar flow is usually slow-
er flow. When the velocity of the flow increases, the flow eventually becomes
turbulent. Turbulent flow is characterized by flow with eddies. Eddies absorb
more energy and create large internal friction, which increases the fluid vis-
cosity. Laminar and turbulent flows are shown in Figures 7-la and 7-lb,
respectively.

(b)
Figure 7-1. (a) Laminar Flow, (b) Turbulent Flow.

For an incompressible fluid continuously flowing through a vessel

(Figure 7-2), the mass of the fluid flowing through section 1 per unit time is
the same as that in section 2. Therefore,
 Flow Transducers

since pi = pz, the equation becomes

A1 x vi = A2 X v2 (2)

-
which shows that the fluid velocity is higher when the cross section is smaller.

Fluid flow

AI -42
Figure 7-2. Fluid Flow Through Pipe with Varying Diameter.

BERNOULLI'S EQUATION

In the early eighteenth century, Daniel Bernoulli developed the

Bernoulli's equation, which expresses the relationships between the velocity,
pressure, and elevation of liquid flowing in a pipe with nonuniform cross sec-
tion and varying height. Consider the setup shown in Figure 7-3. Assuming
the flow is laminar, the fluid is incompressible, and the viscosity is negligible
(i.e., there is no energy loss in the flow), equating the fluid energy at points 1
and 2 produces the following equation, which is called the Bernoulli's equa-
tion.

Pi + -21pvi2 + pghi = P2 + -21 pvi2 + pghn

where p = the density of the fluid, and
g = the acceleration due to gravity.
The equation shows that the quantity P + -1 pv2 + pgh is constant at any
2
point in a lossless fluid flow. Although any fluid flow in real life has certain
energy loss, the Bernoulli's equation provides an estimate of the trade-off
between the pressure, flow, and elevation of the fluid in a vessel.
Bernoulli's equation ignores the friction (called viscosity in fluid analysis)
within a flowing fluid. This friction creates thermal energy, which is loss in
the form of heat. This is why a pump is required (to input energy into the sys-
 Biomedical Device Technology: Principles and Design

Figure 7-3. Fluid Flow in a Pipe.

tem) to move the fluid through a pipe. Viscosity exists in both liquids and
gases. In laminar flow, the fluid layer in immediate contact with the wall of
the pipe is stationary due to the adhesive force between the molecules. The
stationary layer slows down the flow of the layer in contact due to viscosity;
this layer in turns slows down the next layer and so on. Therefore, the veloc-
ity of the fluid varies continuously from zero at the pipe wall to a maximum
velocity at the center of the pipe as shown in Figure 7-4. The viscosity of dif-
ferent fluids can be expressed quantitatively by the coefficient of viscosity 77.
The values of 77 for different fluids stated in Table 7-1 are at 20°C except for
blood and blood plasma, which are measured at 37OC (body temperature).

Zero viscosity Viscosity > 0

(length of arrow indicates the magnitude of flow velocity)

Figure 7-4. Velocity Flow Profile of Fluid in a Pipe.
 Flow Transducers

Table 7-1.
Coefficient of Viscosity for Different Fluids
Fluid Coeficient of viscosity 7 (X la1Pa.s)
-

Water
Ethyl alcohol
Glycerine
Air
Whole blood
Blood plasma
-

For an incompressible fluid, the volume flow rate Qof the fluid under-
going laminar flow in a uniform cylindrical pipe can be expressed by the
Poiseuille's law

where r = the radius of the pipe,

L = the length of the pipe section,
PI and P2 =, respectively, the pressure of the fluid at the beginning and
the end of the section, and
77 = the coefficient of viscosity.
Poiseuille's law shows that the flow rate is directly proportional to the
pressure gradient and to the fourth power of the radius, but inversely pro-
portional to the viscosity of the fluid. From equation (4), one can easily note
that if there is a stenosis in the blood vessel (decreased radius of the vessel),
the blood pressure must increase substantially to maintain the same blood
flow.
When the flow velocity is high, the flow will become turbulent and
Poiseuille's law will become invalid. The average flow rate at a given pres-
sure under turbulent flow is less than that of laminar flow given by equation
(4) since the turbulence creates extra friction in the fluid. The onset of tur-
bulence is often abrupt and is characterized by the Reynolds number Re,
where

where p = the density of the fluid,

6 = the average velocity,
r = the radius of the pipe, and
= the coefficient of viscosity.
Experiments have shown that when Re exceeds about 2,000, the flow
becomes turbulent.
 Biomedical Device Technology: Principles and Desip

Example
The average blood velocity in a vessel of radius 3.0 mm is about 0.4 m/s.
Determine whether the blood flow is laminar or turbulent given that the den-
sity of blood is 1.05 X 10' kg/m3 and the coefficient of viscosity of blood is
4.0 X 10" N.s/m2.

Solution
Using equation (5),the Reynolds number is:

Re = 2 x 1.05 x x 0.4 4 3.0 x 10" = 1,056

4.0 X 10-j
which is less than 2,000
The flow is therefore laminar.

FLOW TRANSDUCERS

There are many flow transducers available. This section describes a num-
ber of common flow transducers that may be found in medical instrumenta-
tion.

Venturi Tube
Figure 7-5 shows the construction of a circular, horizontal Venturi tube
flowmeter. For an incompressible fluid flowing inside a pipe, assuming it is a
frictionless flow, we can apply the Bernoulli's equation at points 1 and 2 of
the fluid in the tube.

as the tube is horizontal, substituting hi = h and rearranging the equa-

tion gives

PI - Pz = J? (vi2- ~ 1 2 )
2 (6)
As the fluid is incompressible, the volume flow of fluid Qpassing through
section 1 and section 2 of the pipe are equal, i.e., Qr = Q2.
But Q= -q where D is the diameter of the pipe and u is the flow
4
 Flow Transducers

Figure 7-5. Venturi Tube.

velocity. Therefore

Substituting equation (7) into equation (6) yields

Therefore, the volume flow rate Qbecomes:

Equation (9) shows that the volume flow rate can be found by measuring
the fluid pressures and pipe diameters at points 1 and 2 of the Venturi tube.

Orifice Plate
A variation of the Venturi tube is the orifice plate (Figure 7--6). The cir-
cular opening of the plate inside the pipe creates a reduction in the diameter
of the fluid flow path downstream of the orifice opening. The differential
pressure A P = PI - P2 is measured to determine the flow velocity v and the
 Biomedical Device Technology: Princ$les and Design

volume flow rate $ Lookup tables of Qagainst A P are usually provided by

Figure 7-6. Orifice Plate Flowmeter.

the manufacturers.
Pitot Tube
A Pitot tube determines the velocity of the fluid by measuring the differ-
ences between the static pressure Ps in the flow and the impact pressure PI.
A Pitot tube has two concentric pipes as shown in Figure 7-7. The inner pipe
has an opening at the outer wall on the side parallel to the direction of fluid
flow and therefore measures the static pressure in the flow. The opening of
the outer pipe is facing the flow and therefore detects the total pressure

Figure 7-7. Pitot Tube.

 Flow Transducers

developed by the moving fluid.

By the Bernoulli's equation, the total pressure is given by Pr = P + pgh
+ -P$,2 and the static pressure is Ps = P + pgh . The differential pressure is
therefore given by:

Compared with the orifice plate or the Venturi tube, a Pitot tube can eas-
ily be installed in the field by drilling a hole on the pipe and inserting the ele-
ment through the hole.

Rotameter
A rotameter (or variable area flowmeter) consists of a float in a uniform-
ly tapered tube as shown in Figure 7-8. An upward flow creates an upward
force on the float. The float will go up or down until the upward force is
equal to the weight of the float. The volume flow rate Qcan be shown to be
proportional to the area of the round gap between the float and the tube. The
volume flow rate is usually calibrated and marked on the side of the tube.
Rotameters are commonly used as reliable and maintenance-free indicators

Fluid flow
Figure 7-8. Rotameter.
 Biomedical Device Technology: Princ$les and Design

of gas flow such as oxygen or medical gases.

Turbine Flowmeters

A turbine (or vane or paddle wheel) installed in a pipe will rotate at a

speed proportional to the fluid flow velocity. A typical example of this class
of flowmeter is a spirometer. Spirometers are used in respiratory gas flow
measurements (Figure 7-9). The vane of a spirometer is made of very light
material supported by bearings to reduce its resistance to the gas flow. Gas
flow causes the vane to rotate at a speed proportional to the flow velocity.
Through a train of gears, the volume of the gas flow through the vane is indi-

- - - Dial indicator
,

- - - - Gear train ,#. #/@

Gas inlet

....
. Vane (rotor)
Cross-sectional view

Figure 7-9. Spirometer.

cated on the readout dial.

Electromagnetic Flowmeters

When a conductive fluid flows across a magnetic field, an electromotive

force (emf) is induced between the two electrodes placed orthogonal to the
magnetic field and the direction of flow (Figure 7-10). The induced emf is
proportional to the average flow velocity of the conductive fluid in the tube.
For a steady flow, factors such as fluid density, viscosity, turbulence, or
Reynolds number, which normally affect other flow measurement methods,
do not affect the output signal. In addition, since none of the components of
the flowmeter is inside the fluid flow path, flow measurement without creat-
ing obstruction to the flow is possible.

Ultrasound Flowmeters
An ultrasound flowmeter using the Doppler principle consists of an ultra-
 Flow Transducers

Voltmeter

Figure 7-10. Electromagnetic Flowmeter.

sound transmitter and a receiver as shown in Figure 7-lla. Particles such as

blood cells in the fluid reflect the ultrasound signal from the transmitter to
the receiver. The frequency shift of the ultrasound signal received by the
receiver is dependent on the velocity of the reflecting particles traveling in
the fluid.
An ultrasound flowmeter using the transit time principle consists of two
ultrasound transmitter and receiver pairs placed in the fluid path is shown in
Figure 7-1 lb. As the sound traveling upstream is slower than that traveling
downstream, the time difference measured between the upstream and down-
stream transit times is used to derive the flow velocity of the fluid in the pipe.
The principles and construction of ultrasound blood flow devices will be cov-
ered in Chapter 27.
Figure 7-12 shows the construction of an ultrasound vortex flowmeter.

U/S Transmitter UIS transmitterlreceiver

Fluid particle
0

Fluid flow Fluid flow

.
UIS receiver
U/S beam
UIS transmitterlreceiver

(a) (b)
Figure 7-11. Ultrasound Flowmeters (a) Doppler, (b) Transit Time.
 Biomedical Dmice Technology: Princ$les and Design

When the flowing fluid hits the wedge located at the center of the vessel, vor-
tices are formed; these vortices create turbulence in the fluid, which can be
detected by an ultrasound transmitter and receiver setup. The number of

L, UIS transmitter

,
Vortex

Fluidflow -+

Wedge
- . / - / / /

UIS receiver
Figure 7-12. Ultrasound Vortex Flowmeter.

vortices formed is proportional to the velocity of the fluid.

Thermal Flowmeters
A thermal flowmeter with a heating element and two temperature sen-
sors (e.g., thermistors) placed at equal distance to the element is shown in
Figure 7-13a. The heater heats the fluid in contact with the heating element.
When there is no flow, the sensors are at the same temperature. The flow of
the fluid causes a temperature imbalance. The flow rate as well as the direc-
tion of flow are determined by the temperature difference between the two

?&\-y:;;, -
.
# .
#
Temperature sensors
,, 8 ,

v-...
Fluid flow -b ..
Heater

Figure 7-13. Thermal Flowmeter.

 Flow Transducers

temperature sensors.
Hot Wire Anemometer
Another type of thermal flowmeter is the anemometer. In a hot wire
anemometer, a wire heated by an electrical current is placed in the fluid flow
path. The fluid flows pass the wire cools it to a lower temperature. The rate
of heat removal from the hot wire is a function of the fluid flow rate. Figure
7-14 shows a hot wire anemometer for respiratory gas flow measurement.
When the gas flow rate becomes higher, it removes heat from the hot fila-
ment at a faster rate. To maintain a constant temperature on the filament, the
flowmeter responses by increasing the filament heating current. The gas flow
velocity is therefore a function of the filament current. To obtain the flow
direction, another hot wire is placed behind a flow diverting pin at the same
level as the first wire. The upstream wire will be cooled more rapidly than

Flow diverting pin

I
Hot filament
\
\
I
I
: Direyional sensing filament
\ /
\ /
\ I /
\ I I
Gas flow

Figure 7-14. Hot Wire Anemometer.

the downstream wire.

A variation of the hot wire anemometer is the hot film anemometer,
which replaces the heated wire with a heated metal film. The hot film pro-
vides a larger contact area between the gas and the heated element to
 Chapter 8

OPTICAL TRANSDUCERS

OBJECTIVES

Identify the frequency band of medical optics in the electromagnetic

spectrum.
Differentiate between photon and thermal optical detectors.
Differentiate between radiometry and photometxy; define and list the
parameters and units of measurements.
Define blackbody, graybody, selective radiators, and color temperature.
Analyze the characteristics of common photosensors.
Study the construction and principles of operation of a charged-couple
device image sensor.

CHAPTER CONTENTS

1. Introduction
2. Thermal and Quantum Events
3. Definitions
4. Types of Photosensing Elements
5. Application Examples

INTRODUCTION

With the rapid advancement of optical instrumentation and communica-

tion systems, optical transducers (also known as photodetectors) have been
 Optical Transducers

undergoing rapid development in the past decade. Many photodetectors

have found applications in medical instrumentation. Photodetectors may be
used to measure energy, flux, intensity, et cetera, of an electromagnetic radi-
ation. Electromagnetic radiation from low (e.g., 2.4 GHz or A = 12.5 cm in
telemetry) to high frequencies (e.g., 3 X lo2' Hz or A = m in Gamma
radiation) are used in various medical applications. Radiometry is the mea-
surement of quantities associated with radiant energy, whereas photometry
is the measurement of quantities associated with visible light. Visible light
falls within a narrow band (A = 380 to 780 nm) in the electromagnetic spec-
trum (Figure 8-1). This chapter covers the principles and construction of
some common optical transducers used in medicine.

,Radio Waves
Microwavesl

1 mm
Infrared
I
750 nm 380 nm
Ultraviolet

10 nm
X Ray
Gamma Ray

0.1 nm
,
0.001 nm
(red) (violet)

Decreasing wavelength (increasing frequency)

Figure 8-1. Electromagnetic Spectrum.

QUANTUM AND THERMAL EVENTS

There are two different mechanisms in radiant energy transduction:

quantum (photon) event and thermal event. In a quantum event, photons are
absorbed by the sensing element, which generates an electrical output
according to the amount and rate of absorption of the light quanta. Quantum
event is governed by the equation:

E = h f = - hc
h
where h = the Planck's constant = 6.625 x lo-" Js,
f = frequency of electromagnetic radiation, Hz,
A = wavelength of electromagnetic radiation, m,
c = speed of light = 3.00 x 10"' ms-I for all frequencies in a vacuum but
lower in other media, e.g., 2.2,5 X 10'" ms-' in water for a 589 nm light source.
As the electrons must gain sufficient energy in order to jump over the
 Biomedical Device Technology: Princ$les and Design

energy gap, a minimum photon energy E m i n is required to initiate the event.

Therefore, radiation with wavelengths longer than Amax (= )- hc will not pro-
Emin
duce a quantum event (define Amax as the maximum wavelength to produce
a quantum event).
In a thermal event, radiant energy is first absorbed by the sensing ele-
ment to produce heat energy; the heat generated in turn causes a change in
an electrical value. Unlike quantum events, thermal events happen at all
wavelengths. The sensitivity of a thermal photosensor also depends on the
wavelength of the incident radiation.
The spectral response of a photosensor is often provided in the device
specifications by the manufacturer. In addition, filters can be used to modify
the spectral response of a sensor.

DEFINITIONS

Thermal and Photon Detectors

Photodetectors can be divided into two classes: thermal detectors and
photon detectors. The sensing element of a thermal detector first converts
the electromagnetic radiation to heat and then to an electrical value. This
change in electrical value is often proportional to the received energy. That
is, they are using the thermal event of transduction. Photon detection is
achieved by quantum events that stimulate the sensing element to produce
an electrical signal proportional to the rate of absorption of light photons.

Radiometry and Photometry

Radiometry is the measurement of quantities associated with radiant
energy. The units of measurements and definitions of some radiometric para-
meters are listed below:
Radiant Energy-Energy traveling in the form of electromagnetic
waves.
Radiant Density-Radiant energy per unit volume.
Radiant Flux-The time rate of flow of any parts of the radiant energy
spectrum.
Radiant Flux Density at a surface-The quotient of radiant flux to the
area at that surface. Radiant emittance (M) refers to radiant flux den-
sity leaving a surface while irradiance (E) refers to radiant flux densi-
ty incident on a surface.
 Optical Transducers

Radiant Intensity-The radiant flu from the source per unit solid
angle.
Radiance-The radiance in a direction at a point on a surface is the
quotient of the radiant intensity leaving, passing through, or arriving
at the surface to the area of the orthogonal projection of the surface on
a plane perpendicular to the given direction.
In photometry or lighting engineering applications when onIy the visible
part (for human eyes, A = 380 to 780 nm) of the electromagnetic spectrum is
of interest, the term "radiantn is changed to Kluminousnfor the measurement
quantities. For example:
Luminous Energy-radiant energy of the electromagnetic radiation in
the visible spectrum (from 380 to 780 nm).
Luminous Flux-The time rate of flow of the luminous part (380 to 780
nm) of the radiant energy spectrum.
The quantities, symbols, and SI units of these parameters are tabulated
in Table 8-1.

Table 8-1.
- --
Standard Units of Radiometry and Photometry
- - - - - - - - -

Radiomatric Parameter SI unit Photometric Parameter SI Unit Equation

Radiant Energy (Q) J Luminous Energy (Q) 1m.s

Radiant Density (R) Luminous Density (R)

Radiant Flux (@) Luminous Flux (@)

Radiant Flux Density Luminous Flux Density

at a Surface at a Surface
Radiant Emittance (M) Luminous Emittance (M)

Irradiance (E) Illuminance (E)

Radiant Intensity (I) Luminous Intensity (I)

Radiance (L) w/(sr.m2) Luminance (L) nit

Blackbody, Graybody, and Selective Radiators

The radiant energy from a practical source is often described by com-
paring it with a blackbody radiator. A blackbody is defined as an entity that
absorbs all incident radiation, with no transmission or reflection. A black-
body also radiates the most power at any given wavelength than any other
 Biomedical Device Technology: Principles and Design

source operating at the same temperature. The ratio of the output power of
a radiator at wavelength A to that of a blackbody at the same temperature
and the same wavelength is known as the spectral emissivity 6 (A) of the radi-
ator. When the spectral emissivity of a radiator is constant for all wave-
lengths, it is called a graybody. No known radiator has a constant spectral
emissivity over the entire electromagnetic spectrum. However, some materi-
als exhibit near graybody characteristics within a certain range of wave-
lengths (e.g., a carbon filament in the visible region). Non-graybody radiators
are called selective radiators. The emissivity of a selective radiator is differ-
ent at different wavelengths.
The color temperature of a selective radiator is equal to that temperature
at which a blackbody must be operated in order that its output characteristic
be the closest possible to a perfect color match with the output of the selec-
tive radiator (Figure 8-2). Note that color matching does not imply equal
radiant output.

Radiant flux
per unit
wavelength A

r
Wavelength
Figure 8-2. Radiant Curves for (a) Blackbody, (b) Graybody,
and (c) Selective Radiator at the Same Color Temperature.

Although the match is never perfect, color temperature values are used
to represent spectral distribution of light sources. For example, sunlight and
the output of a tungsten filament incandescent lamp have color temperatures
of 6500 K and 3000 K, respectively. Table 8-2 shows the visible color corre-
sponds to the absolute temperature (K) of a blackbody radiator.
 Optical Transducers

Table 8-2.
-
Color Temperature
- - -
-
of--a Blackbody
- - -
- -
Radiator
- - -

- -
Temperature (K)
- -
Color - -

800-900 Red
3000 Yellow
5000 White
8000-10000 Blue
60000- 100000
- - - - -
Brilliant sky blue
-- - - -

TYPES OF PHOTOSENSING ELEMENTS

Some examples of photosensors are discussed in this section. Photoresis-

tive sensors, thermopiles, and pyroelectric sensors are primarily thermal
event detectors, while photoconductors, photoemissive sensors, photodi-
odes, phototransistors, and charge-coupled devices are quantum event detec-
tors.

Photoresistive Sensors
For a semiconductor material, the electron and hole mobility, and hence
its resistivity, varies with temperature. This property of temperature-depen-
dent resistivity of semiconductor material can be used to detect thermal ener-
gy from a radiant source. The resistivity increases with temperature for light-
ly doped silicon but decreases at high doping level.

Thermopiles
A thermopile is made up of a number of thermocouples connected in
series (Figure 8-3). Radiant energy is first converted to heat, creating a dif-
ferential temperature between the hot and cold junctions of the thermocou-
ples. Each thermocouple will generate a small voltage (on the order of pVs)
according to this temperature difference. The output of the thermopile is the
summation of all the voltages from the thermocouples in the cell. The sensi-
tivity of the thermopile is proportional to the number of thermocouple ele-
ments in series. Miniature thermopiles, such as Si/Al thermopile, are fabri-
cated from microelectronic technology.
 Biomedical Device Technology: Principles and Design

Illumination ---------,
-.--.
*.
•.. 1 v

L--------.

T
Hot
T
Cold
junctions junctions
Figure 8-3. Thermopile Composed of Four Thermocouples.

Pyroelectric Sensors
In a pyroelectric sensor, a conductive material is deposited on the oppo-
site surfaces of a slice of ferroelectric material, such as triglycine sulfate
(TGS).The conductor on the sensing side is transparent to the light source
to be measured. The ferroelectric material absorbs radiation and converts it
to heat. The resulting rise in temperature changes the polarization of the
material. The current I flowing through the external resistor connected
across the two conductive surfaces is proportional to the rate of change of
temperature of the sensor. Pyroelectric sensors with cooling can detect radi-
ant power down to lo-" W with A- of up to 100 pm.

Photoconductive Sensors
In a semiconductor, electrons can be raised from the valence band to the
conduction band by absorbing energy from light photons. The presence of
these photon-induced electrons increases the conductivity of the semicon-
ductoi material. In order to raise the electron from the valence band to the
conduction band, it must absorb enough energy from the photon to over-
come the band gap energy Ec. Therefore, to create the photoconductive
effect, the energy of the light photon should be greater than Ec and hence
the wavelength of the radiation must be smaller than Amax given by equation
(2):

For high-precision applications, the sensing element often must be

cooled to reduce noise from thermionic electrons, which decrease the signal
to noise ratio of the sensor.
 Optical Transducers

Photoemissive Sensors
The phenomenon that electrons are liberated to the free space from the
surface of a material when excited by light photons is called photoelectric
effect. For high-efficiency conversion, the potential barrier or work function
Eo of the material must be much smaller than the photon energy. The effi-
ciency or quantum yield of the sensor is defined as the ratio of the number
of emitted electrons to the number of absorbed photons. Materials with low
Eo, such as NaKCsSb or cesium oxide on GaAs substrate, have high quan-
tum yields and therefore are good materials for photocathodes (photoemis-
sive elements). In a simple photoelectric tube (Figure 8-4), under light illu-
mination, electrons are liberated from the photocathode and conducted
through the external circuit. If the wavelength is shorter than the work func-
tion, the photoelectric current produced in a vacuum photoelectric tube
changes linearly with the level of illumination. In a gas-filled tube (e.g., filled
with low-pressure argon), collisions of the electrons with gas atoms may pro-
duce secondary electrons emissions, resulting in higher sensitivity photode-
tection.

Photo cathode

Figure 8-4. Photoelectric Tube.

Photodiodes
A diode is constructed of an n-type semiconductor (e.g., phosphorus
doped silicon) in contact with a p-type semiconductor (e.g., boron doped sil-
icon). Under normal conditions, electrons readily break away from the impu-
rity in the n-type material to become free electrons. In the p-type semicon-
ductor, mobile holes are created instead. Despite the presence of mobile
electrons in the n-type and mobile holes in the p-type semiconductor, the
 Biomedical Device Technology: F'rinciples and Design

individual semiconductors are both neutral in charge (i.e., the number of

electrons is the same as the number of protons). However, when the two
types of semiconductors are put together, the electrons from the n-type semi-
conductor will migrate across to fill the holes in the p-type semiconductor,
forming a P-N junction (Figure 8-5). This migration of electrons also creates
a net positive charge in the n-type and a negative charge in the p-type semi-
conductor. These charge layers create a net electric field and will eventually
prevent further electrons from traveling across the junction. The electric field
at the junction acts as a diode, creating a barrier for electrons to move from
the n-type to p-type but not the other way around.

P-type
Semiconductor

N-type
Semiconductor

Figure 8-5. Semiconductor P-NJunction.

When light is incident upon the P-N junction of a semiconductor, a pho-

ton with sufficient energy will free a number of electrons from the atom. The
number of free electrons (or accumulation of charges) can be found by mea-
suring the voltage across the photodiode (photovoltaic mode) or by measur-
ing the reverse bias current of the photodiode (photoconductive mode). The
voltage-current characteristic of a typical photodiode and its equivalent cir-
cuit are shown in Figure 8-6 where v is the voltage across the diode and i is
the diode current. The current source represents the current produced when
the junction is exposed to radiation. Notice that when the output is short-cir-
cuited (i.e., v = O), the short-circuit current is negative and its magnitude
depends on the wavelength and the intensity of the incident illumination.
In the photoconductive mode of operation, a reverse bias voltage is
applied to the photodiode as shown in Figure 8-7 (i.e., v is negative). The
magnitude of this reverse bias current i will increase with the intensity I of
the illumination. Although the reverse bias current is almost the same for a
constant illumination, a higher reverse bias voltage will produce a faster
response time. However, dark current (noise) increases with the magnitude
of the applied bias voltage. Diodes in photoconductive modes operate in
quadrant A of the photodiode characteristic curve (Figure 8-6).
 Optical Transducers

Figure 8-6. Current-Voltage Characteristic of Photodiode

with Illumination Intensity Io<Il<h.

Photocurrent

P-type
Semiconductor ......................
A- Reverse
bias

T
N-type voltage
Semiconductor

Figure 8-7. Photovoltaic Cell.

In photovoltaic mode of operation, no bias voltage is required. The free

electrons created in the p-type semiconductor, being moved by the positive
electric field, will cross the junction into the n-type semiconductor. These
electrons will accumulate at the junction and create a potential difference
(photovoltaic) between the p-type and n-type semiconductors. A photocur-
rent will flow if an external path is established between the semiconductors
(Figure 8-8). This is often referred to as a solar cell. The magnitude of the
photocurrent increases with the light intensity. Since dark current is a func-
tion of the bias magnitude, photovoltaic mode is ideal for low signal level
detection. However, it is less suitable for high-frequency application because
of its slow respond time. Diodes in photovoltaic modes operate in quadrant
 Biomedical Device Technology: Principles and Design

B of the photodiode characteristic curve (Figure 8-6). Other than silicon,

semiconductor materials such as gallium arsenide, and copper indium dise-
lenide are common materials for photodiodes.

,
Light photon
\
\ r-l
I
Photocurrent
External

P-type
Semiconductor
A' e-,
1
u
Figure 8-8. Photovoltaic Cell (Solar Cell).

Phototransistors
A phototransistor can provide amplification of the photocurrent within
the sensing element. The basic construction of the phototransistor is similar
to that of a bipolar transistor except that the base normally has no connec-
tion and is exposed to the illumination being measured. For the phototran-
sistor connection shown in Figure 8-9, the emitter current is given by:
i~ = ( i + +
~ i~)(hfe 1)
where i~ = light-induced base current,
i~ = reverse leakage current, and
hj = forward current transfer ratio.
If i~ >> i~ and hj >> 7, the above equation can be simplified to
i~ = i~hfe (3)
Equation 3 shows a linear high gain characteristic, which converts luminous
intensity to a relatively large emitter current.

Charge-Coupled Device (CCD)

Charge-coupled devices (CCDs) are often used in digital image acquisi-
tions. The sensing elements of a CCD are MOS capacitors. Figure 8-10 illus-
 Optical Transducers

--.LJ
Light photons

Figure 8-9. Phototransistor Circuit.

trates the cross section of a buried channel MOS capacitor. A typical buried
channel MOS capacitor has a n-type semiconductor layer (about 1 p,m)
above a p-type semiconductor substrate with an insulation formed by grow-
ing a thin oxide layer (about 0.1 p,m) on top of the n-type layer. A conduc-
tive layer (metal or heavily doped semiconductor) is then deposited on top
of the oxide to serve as the metal gate.

Metalgate -----
Oxide layer ----

1
N-type layer

--------------- (----
P-N junction

P-type substrate ----7

Figure 8-10. Buried Channel MOS Capacitor.

When light photons are allowed to incident on the P-N junction, elec-
tron-hole pairs are formed. Similar to a photodiode, the electrons will
migrate to the n-type side of the junction and will be trapped in the "buried
channel" (Figure 8-11). To create separation between adjacent pixels, a p-
type stop region on each side of the metal gate is formed to confine the
charges under the gate. The amount of trapped charge is proportional to the
number of incident light photons.
A CCD consists of an array of these individual elements (pixels) built on
a single substrate. A 256 X 256 array CCD contains 2"' number of MOS
capacitor elements. To understand the operation of CCDs, we can represent
 Biomedical Device Ethnology: Princ$les and Design

P-type stop region

Metal gate

Oxide layer Trapped charges

/

N-type layer
....................... P-N junct~on

P-type substrate

Figure 8-1 1. A CCD Pixel.

a MOS capacitor element by a bucket and the amount of trapped charges by

the level of water in the bucket. When light of nonuniform intensity incidents
on the CCD array for a predetermined duration of time (usually controlled
by a timer-controlled shutter), the amount of charges created and trapped in
each MOS capacitor is proportional to the number of light photons received
by the capacitor. Figure 8-12 shows a 1 X 3 CCD linear array at the top and
its water bucket analogy at the bottom. A larger amount of light photons cre-
ates a brighter CCD pixel, which is represented by a higher water level in
the bucket.

CCD pixel
Pixel 1 Pixel 2 Pixel 3

v2 v3 I Vl v2 v3 I

Figure 8-13. CCD Charge Transfer Process.

 Biomedical Device Technology: Principles and Design

The process of reading the amount of charge in each pixel involves mov-
ing the charges from the site of collection to a charge-detecting circuit locat-
ed at one end of the linear array. Reading out the charges in a CCD array is
a sequential process. This process is illustrated in Figure 8-13.
In general, each pixel has three gates; each gate is connected to a time-
controlled bias voltage K, Vz, and fi as shown (upper diagram of Fig. 8-13).
Consider the charges stored in pixel number 2. At time Ti, fi becomes pos-
itive. As a result, some charges below the Irz electrode migrate to the region
under fi. At time T2, the Vz electrode is turned off; the rest of the charges
under the Irz electrode have now all migrated to the region under fi. Similar
charge shifts happen at time T3, E, T5, and T6. After time T6 (one shift cycle),
the charges under pixel 2 has been shifted to pixel 3. Note that all charge
packets in the array move (or shift) simultaneously one pixel to the right. For
an N-pixel linear array, N shift cycles are required to read out the entire
array.
Figure 8-14 shows the physical layout of a 3 X 2 (6 pixels) CCD array.
The pixels are represented by the rectangular boxes. In this particular con-

Gates for
pixels

l stops
 Optical Transducers

figuration, reading out the charges is achieved by shifting the charges in the
vertical direction.
CCDs used in imaging are usually in square or rectangular arrays. An N
X M array (Figure 8-15) can be considered to be made up of M linear arrays,
each with N pixel elements. Reading out the array requires simultaneously
shifting all rows of charge packets one pixel downward toward the serial reg-
ister. The charge packets are then transferred from the serial register to the
output amplifier one row at a time.

Figure 8-15. Two-Dimension CCD Array.

The unwanted charge accumulated in CCD pixels due to thermally gen-

erated electrons instead of light photons is called the dark current. Dark cur-
rent contributes to noisy picture, especially when the input light level is low
(i.e., low signal to noise ratio). As this noise is due to thermal effects, dark
current can be reduced by cooling the CCD to a lower temperature.
Quantum efficiency (QE) measures the efficiency of incident photon detec-
tion. It is defined as the ratio of the number of detected electrons Ndivided
by the number of incident photons M times the expected number of elec-
trons R generated from each photon.
 Biomedical Device Technology: Principles and Design

APPLICATION EXAMPLES

There are many applications of optical sensors in medicine. A few exam-

ples are:
LED-photodiode pair in infusion pump fluid drop sensors
LED and photosensors in pulse oximeters
Thermopiles in tympanic thermometers
Photodetectors in capnometers
CCD cameras in endoscopic video systems
Ambient light sensors in monitor screen dimmers
Flat panel detectors in digital x-ray systems
 Chapter 9

ELECTROCHEMICAL TRANSDUCERS

Differentiate galvanic and electrolytic electrochemical cells.

Describe the construction of a galvanic cell and analyze the function of
each component with respect to the electrochemical reactions.
Define standard half-cell potential and compute cell potential under
nonstandard conditions.
Explain how galvanic cells can be used as electrolyte analyzers.
Analyze the constructions and the electrochemical reactions of refer-
ence electrodes including hydrogen, calomel, and Ag/AgCl standard
electrodes.
Analyze the constructions and electrochemical reactions of ion selective
electrodes including pH, pK, and pCOz.
Define amperometry and potentiometry.
State the principles of enzyme biosensors.
Describe the principles of batteries and fuel cells.

CHAPTER CONTENTS

1. Introduction
2. Electrochemistry
3. Reference Electrodes
4. Ion Selective Electrodes
5. Biosensors (Enzyme Sensors)
6. Batteries and Fuel Cells
 Biomedical Device Technology: Princ@lesand Design

INTRODUCTION

Electrochemistry plays an important role in understanding as well as

measuring biopotential signals. This chapter covers the principles of galvan-
ic cells and electrode potentials. Reference electrodes including hydrogen,
Ag/AgCl, and calomel electrodes are studied. Application examples such as
K+, pH, pOz, and pCO2 electrodes are discussed to illustrate the principles
of ion selective electrodes. Enzyme sensors for glucose and cholesterol mea-
surement are introduced to illustrate this growing field of biosensors. The
principles of energy storage cells and fuel cells are also discussed.

ELECTROCHEMISTRY

Electrochemistry is the study of the interconversion processes concern-

ing the relationship between electrical energy and chemical changes. An
electrochemical cell is a device that permits the interconversion of chemical
and electrical energy. There are two types of electrochemical cells: galvanic
and electrolytic cells. In a galvanic cell, also known as a voltaic cell, chemi-
cal energy is converted to electrical energy. The reverse happens in an elec-
trolytic cell, where electrical energy is converted to chemical energy. In the
context of medical instrumentation, we consider only galvanic cells in this
chapter.

Galvanic (Voltaic) Cells

In a galvanic cell, the electrical energy produced is the result of a spon-
taneous redox reaction between the substances in the cell. Redox reaction is
a chemical reaction involving oxidation (losing electrons) and reduction
(gaining electrons). Consider the simple redox reaction by placing a zinc rod
into a solution of copper sulfate. One would immediately notice a black dull
deposit forming on the shiny zinc surface. The chemical equation of this
spontaneous reaction is:
Zn(s) + Cu" + Zn" + Cu(s)
In this chemical reaction, copper ions from the solution are reduced to
fine particles of copper metal, which grow to form a spongy layer on top of
the zinc metal. The blue color of the copper sulfate solution gradually
becomes paler (from its blue color), indicating that hydrated copper ions are
used up in the reaction. The redox equation can be separated into an oxida-
 Electrochemical Transducers

tion equation and a reduction equation as shown below. Each of these is

called a half-reaction.
Zn(s) + Zn2++ 2e- (oxidation)
Cu2++ 2e- + Cu(s) (reduction)
The free-energy change of the overall redox reaction is -212 kJ when the
reactants and products are in their standard states (i.e., pure metal and 1 M
ionic concentration). This large negative free-energy indicates a strong ten-
dency for electrons to be transferred from the zinc metal to the copper ions.
The free-energy change in a reaction depends only on the nature and state
of the reactants and products and not on how the process takes place.
Consider the setup in Figure 9-1, where the zinc rod is separated from
the copper sulfate solution. In the setup, the zinc rod is immersed in a solu-
tion of zinc sulfate and a copper rod is immersed in a solution of copper sul-
fate. The two compartments of the cell are separated by a porous partition
such as a piece of porcelain or clay. The zinc and copper rods become elec-
trodes to provide surfaces at which oxidation and reduction half-cell reac-
tions can take place.

e-
.-.---.--
b 1
Copper electrode
Zinc electrode I b I (cathode)
(anode)

ZnS04 solution CuSO, solution

Figure 9-1. Galvanic Cell with Porous Partition.

If the zinc and copper electrodes are connected by an external conduc-

tor, electrons will leave the zinc metal and travel through this external con-
ductor to the copper electrode. These electrons reduce the copper ions in the
solution adjacent to the copper electrode to form atomic copper, which
 Biomedical Deuice Technology: Pfinc$les and Design

deposits on the surface of the copper electrode. In this galvanic cell, the
anode (the electrode at which oxidation takes place) is the zinc electrode and
the cathode (the electrode at which reduction takes place) is the copper elec-
trode. The positive ions (in this case Zn") which migrate toward the cathode,
are called cations, whereas the negative ions (So?), which migrate toward
the anode, are called anions.
When the external circuit is broken and a voltmeter of high-input imped-
ance is connected across the cathode and anode, the voltmeter reading rep-
resents the electrical potential difference of the galvanic cell. The cell voltage
depends on the type of metal and its electrolyte concentration in each of the
two half-reaction compartments. The potential difference measured between
the cathode and the anode for this zinc-copper cell at 1 M ionic concentra-
tion and 25OC is 1.10 V (note that the cathode is positive and the anode is
negative).
The purpose of the porous partition is to prevent direct transfer of elec-
trons in the solution from the zinc metal to the copper ions. Without the
porous partition, there will be no electron flow in the external circuit as the
copper ions are able to migrate in the solution to the zinc electrode and cap-
ture the electrons directly from the anode. The porous partition in Figure 9-1
can be replaced by a "salt bridge" as shown in Figure 9-2. A salt bridge is a
tube filled with a conductive solution such as potassium chloride (KC]). In
the salt bridge, K+migrates toward the cathode and C1- toward the anode. A
salt bridge provides physical separation between the galvanic cell compart-
ments and establishes electrical continuity within the cell. In addition, it
reduces the liquid-junction potential. Liquid-junction potential is a voltage

e- ------*--
Salt bridge (KC1 aq)

ZnS04 solution Cotton plug CuS04 solution

Figure 9-2. Galvanic Cell with Salt Bridge.
 Electrochemical Tranrducers

produced where two dissimilar solutions are in contact and when the rates of
migration of the cations and anions are not the same across the contact
region (or the junction). The ions in the salt bridge are chosen such that
cations and anions migrate across the junction at almost equal rates, thus
minimizing the liquid-junction potential. Calculation of the cell voltage is
simplified if no liquid-junction potential is present.
The Zn-Cu galvanic cell we have discussed so far in which electrons must
travel through an external circuit before reaching the cathode compartment
is called a Daniell cell. The galvanic cells described in Figures 9-1 and 9-2
are both Daniell cells. Galvanic cells can be depicted by a line notation
called a cell diagram. The cell diagram of the Daniell cell is:
I I
Zn(s) ZnSOl(aq) CuSO4(aq) Cu(s) I
In this cell diagram, the anode is at the left and the vertical lines represent
the junctions. When the salt bridge is present, the junction is represented by
a double vertical line:

This can be further simplified by showing only the reacting ions in the solu-
tion phases:
I
Zn(s) Zn" 11 Cu2' I CU(S)
Standard Electrode Potentials
Consider the two half-reactions of the Daniel cell:
(oxidation)
(reduction)
If we reverse the first half-reaction, it becomes a reduction reaction:
Zn" + 2e- + Zn(s) (reduction)
It is not possible to measure the absolute potential of a single electrode
(or half-cell) a . all measuring devices can measure only the difference in
potential. However, we can use a standard reference electrode to establish
the half-cell potentials of different electrodes. The standard electrode poten-
tials are measured against a hydrogen electrode under standard conditions,
that is, all concentrations are 1 M, partial pressure of gases of 1 atmosphere,
and temperature at 25OC. The potential of this standard reference hydrogen
electrode is given a value of zero volt. Figure 9-3 shows the setup of a stan-
dard reference hydrogen electrode used to measure the standard electrode
potential of a metal Me.
 Biomedical Device Technology: Princ$les and Design

Figure 9-3. Measurement of Standard Electrode Potential.

By convention, the half-cell reaction is written as a reduction reaction

and the potential measured is denoted as the standard reduction potential E":
Men++ ne- + Me(s) (reduction)
Table 9-1 lists the standard electrode potentials of some half-cell reactions.

Table 9-1.
Standard Electrode Potentials
- - -

Half-Reaction (Reduction) E0
--
o?)
Lif + e- + Li(s) -3.05
Na+ + e- + Na(s) -2.71
Al3++ 3e- + Al(s) -1.66
Zn2++ 2e- + Zn(s) -0.76
Fe'" + 2e- + Fe(s) -0.44
2H+ + 2e- + Hz(g) 0.00
AgCl(s) + e- + Ag(s) + C1- +0.22
Hg~Clz(s)+ 2e- + 2Hg(l) + 2C1- +0.27
Cu" + 2e- + Cu(s) +0.34
Fe3' + e- + Fe"' +0.77
Ag' + e- + Ag(s) +0.80
Oz(g) + 4H+ + HzO(1) +1.23
Cln(g) + 2e- + 261- - -
+ 1.36

Note that the sign of the standard potential may be positive or negative.
A more negative standard potential implies higher reducing strength of the
reaction. The standard oxidation potential has the same magnitude but
 Electrochemical Transducers

opposite polarity to the standard reduction potential. For example, the oxi-
dation potential of the copper half-reaction is -0.34 V:
Cu(s) + Cu" + 2e-
The potential of a galvanic cell E0ce1lis the sum of the standard reduction
potential for the reaction at the cathode EOcathodeand the standard oxidation
potential for the reaction at the anode (-Enanode).Therefore, in the case of the
Daniel1 cell,

Example 7
Find the standard voltage produced by the cell
I
Ag(s) AgCl C lI 11 Cu" I Cu(s)
Solution
The separate half-reactions of the cell are:
Ag(s) + C1- + AgCl(s) + e- EOox= -Eu = -0.22 V
Cu2++ 2e- + Cu(s) E" = +0.34 V
For the galvanic cell:
+
2Ag(s) + 2C1- Cu2+ 2AgCl(s) + Cu(s)
E"C~II= +0.34 - 0.22 = +0.12 V

Example 2
Find the standard voltage produced by the cell
I 1) I
Pt(s) Fez+,Fe3+ Cl- C12(g) Pt(s) I
Solution
The separate half-reactions of the cell are:
Fe" + Few + ee Enox = -Eo = -0.77 V
C12(g) + 2e- + 2C1- En = +1.36 V
 Biomedical Device Technology: Princ$les and Design

For the galvanic cell:

Nernst Equation
Consider a hypothetical half-cell reaction:

When it is under nonstandard condition (i.e., concentration other than 1 M

and T is not equal to 25OC), the cell voltage can be quantitatively estimated
by the Nernst equation:
E = Eo - 2.303RT
nF lo&
where R = the gas constant = 8.314J K-' mol-',
T = the temperature in Kelvin,
F = the Faraday constant = 96,485 C mol-', and
[Redj
Q= the reaction quotient = where concentrations are in mol-..
[Oxla
At 25OC, the Nernst equation can be simplified to:

Note that the concentration for solid and liquid is given a value of 1 in the
equation.

Example 3
a) Find the half-cell potential of a copper electrode immersed in a 0.010 M
Cu" solution at 25OC.
b) If the above half-cell is connected to a standard zinc half-cell, what would
be the cell potential?

Solution
a) The half cell reaction is

Cu" + 2e- + Cu(s)

 Electrochemical Transducers

At standard condition, E0 = +0,34 V.

E =
0.0592
0.34 - - ol :0 0592
I' = 0.34 g- 100 = 0.34 - 0.0592 = +0.28 V
2 log 0.010' 2
b) Since Eanode = Enzinc = -0.76 V, the cell potential is
E c e ~= Ecathode + (-Eanode) = +0.28 + 0.76 = + 1.04 V
The cell potential is + 1.04 V (instead of + 1.10 V under standard condition).
Example 4
When the copper half-cell in Example 3 is connected to a nonstandard zinc
half-cell at 25OC, the cell potential is measured to be +0.98 V. What is the
Zn" concentration in the solution?

Solution
The zinc half-cell reaction (E" = -0.76 V) is
Zn" + 2e- + Zn(s)
Since Ece11 = Ecathode + (-Eanode)
Ezinc = Ecathode - Ecell = 0.28 - 0.98 = -0.70 V
and

Using the Nernst equation

1
2.027 = log -
X

The concentration of Zn" in the solution is 0.0094 M.

We have shown in Example 4 that the concentration of a cation in a solu-
tion can be measured by using a reference electrode and a metal indicator
electrode of the same cation. Consider the example of a silver indicator elec-
 Biomedical Device Technology: Principles and Design

trode to find the concentration of Ag+in a solution. The half-reaction and the
standard potential are:
Ag++ e- Ag(s) E" = +0.80 V
Using the Nernst equation, the half-cell potential is:

E = +0.80 -
0.0592
- 1 = +0.80 - 0.0592 log - 1
1 log [Ag+] 145'1
From the above equation, if the half-cell potential E is known, then the con-
centration of Ag+ can be calculated.
Note that if more than one ion are reduced (or oxidized) at the same
time, such a metal indicator electrode will fail to provide the correct mea-
surement. For example, a copper electrode will measure both Cu2+and Ag+
ions in the solution. Furthermore, many metals react with dissolved oxygen
from the air and therefore deaerated solution must be used.
A metal electrode can also be used to measure the concentration of an
anion if the cations of the metal form a precipitate with the anions. Consider
a silver electrode in a saturated solution of AgCl.
AgCl(s) H Ag+(aq)+ C1-(aq)
The solubility product Kip is given by

Substituting [Ag+]into the Nernst equation gives

The concentration of C1- can therefore be calculated if the half-cell

potential E is known.

REFERENCE ELECTRODES

Example 4 shows that one can use the Nernst equation to determine the
concentration of an analyte (in this case Zn2+ion concentration) in the solu-
tion of a half-cell by measuring the potential difference of the cell when the
potential of the other half-cell is known. A typical cell used in potentiomet-
ric analysis is denoted by:
Reference Electrode 11 Analyte ( Indicator Electrode
The cell voltage is the potential difference between the indicator elec-
 Electrochemical Transducers

trode and the reference electrode. An indicator electrode is a half-cell whose

potential varies in a known way with the concentration of the analyte, where-
as a reference electrode is a half-cell with a known constant potential that is
independent of the analyte. In addition to the standard hydrogen electrode
mentioned above, two other commonly used reference electrodes are dis-
cussed in this section. They are the silver/silver chloride (Ag/AgCl) electrode
and the calomel electrode.

Hydrogen Reference Electrode

The hydrogen electrode is a gas-ion electrode. A gas-ion electrode uses a
gas in contact with its anion or cation in a solution. The gas is bubbled into
the solution and electrical contact is made by means of a piece of inert metal,
usually platinum. Figure 9-4 shows the construction of a hydrogen electrode.
The cell diagram and electrode half-reaction of this hydrogen half-cell are:
pt(s) I Hz(g) I H' 11 +
2H' 2e- + Hz(g)
Under standard condition (1 atm. and 25OC), the hydrogen half-cell

Connection wire
bridge

Figure 9-4. Hydrogen Reference Electrode.

potential is assigned a value of 0 V.

Silver/Silver Chloride Reference Electrode
A silver/silver chloride (Ag/AgCl) electrode consists of a piece of silver
 Biomedical Device Technology: Principles and Design

coated with AgCl immersed in a solution of potassium chloride (KC1) satu-

rated with AgCl. Figure 9-5 shows a reference Ag/AgCl electrode (dotted
line) replacing the hydrogen electrode in Figure 9-3 to measure the standard
electrode potential of a metal Me or the ionic concentration of the metal in
the solution. The cell diagram of this half-cell is:

The half-cell reaction and the standard reduction potential E", respectively,
are:

AgCi covered --
Ag electrode

AGIAgCI
reference
electrode w

r.----.--..---..-.-----------------------:

Figure 9-5. Measurement Using a Ag/AgCl Reference Electrode.

AgCl(s) + e- + Ag(s) + C1- E"= 0.22 V

Applying the Nernst equation to this half-cell reaction at 25OC, the half-
cell potential of the Ag/AgCl electrode is

This potential depends on the concentration of the chloride ion in the

KC1 solution. Instead of using a standard 1 M solution, in order to maintain
a constant concentration (to keep E constant), a saturated solution of KC1
(approx. 4.6 M) is used. With a saturated KC1 solution and at 25OC, the half-
cell voltage becomes 0.20 V.
Figure 9-6 shows a typical commercial Ag/AgCl reference electrode.
The salt bridge is replaced by a porous plug at the base of the electrode to
allow passage of ions and completion of the electrical circuit. The air vent
allows the electrolyte to drain slowly through the porous plug. The presence
of some solid KC1 and AgCl at the bottom maintains a solution saturated
 Electrochemical Transducers

Silver wire Air vent

Saturated solution
of KC1 and AgCl
AgIAgCI electrode

Porous plug

Solid KC1 and AgCl

Figure 9-6. Ag/AgCl Reference Electrode.

with KC1 and AgC1.

Calomel Reference Electrode
A calomel electrode consists of a platinum electrode in contact with mer-
cury, mercury chloride (also known as calomel), and a KC1 solution of
known concentration. The half-cell is denoted by the cell diagram:

where x = the molar concentration of KC1 in the solution.

The half-reaction and standard reduction potential (1M C 1 ions) is given
by:
+
Hg~Clz(s) 2e- 2Hg(l) + 2C1- E0= 0.27 V
In practice, saturated KC1 (4.6 M) is often used. The advantage is that the
concentration, and therefore the potential, will not change even when some
of the solution evaporates. At 25OC, a saturated calomel electrode (SCE) has
a potential of 0.24 V. Figure 9-7 shows the construction of a laboratory
calomel electrode. Similar to the Ag/AgCl reference electrode, the salt
bridge is replaced by a porous plug in a commercial calomel reference elec-
trode (Figure 9-8).
 Biomedical Device Technology: Principles and Design

I
Connection wire

Saturated KC1

Figure 9-7. Laboratory Calomel Reference Electrode.

Air vent

Inner tube

Porous plug

Small hole

Figure 9-8. Commercial Calomel Reference Electrode.

ION SELE(;TIVE ELECJXODES

An ion selective electrode (ISE) is an indicator electrode that produces a

voltage when it is in contact with a solution of a particular ion. In general,
this is achieved by using a membrane that is selective for the ion being ana-
lyzed. An ion selective electrode is also known as a membrane indicator elec-
 Electrochemical Transducers

trode.
pK Electrode
An example of an ISE is a K+ electrode or pK electrode. The ion selec-
tive membrane of a pK electrode is a hydrophobic synthetic material con-
taining ionophores such as valinomycin. The membrane is impermeable to
H+,OH-, K+,C1-, et cetera. Ionophores are antibiotics produced by bacteria,
which are used to facilitate the movement of cations across the synthetic
membrane. Valinomycin (an ionophore) can bind K+ tightly but has a 1,000
times lower affinity for Na+. The valinomycin-K+complex can readily pass
through the membrane, from a solution of high K+concentration to a solu-
tion of low K+ concentration. In the case of a solution of 0.1 M KC1 and a
solution of 0.01 M KC1 separated by the membrane, potassium ions (K+)are
transported via the valinomycin-K+ complex from the high concentration
solution into the low concentration solution. However, as C1- ions cannot
pass through, a slight positive potential will develop on the low concentra-
tion side of the membrane with respect to the high concentration side. This
potential eventually stops the net transfer of the K+ ions across the mem-
brane. The potential across the membrane is the difference in potential El
and Ez at the surfaces of membrane due to different concentration of K+ at
both sides of the membrane. At 25OC this membrane potential is given by:

= E l - & = - 0.0592 log 0.01

= 0.059 log -= -0.059 v
n [Pa] 0.1

Calomel Analyte K+
reference
electrode

Figure 9-9. Measurement of K+Concentration Using an ISE.

 Biomedical Device Technology: Principles and Design

where n is the charge on the ion; for K+, n = 1.

Figure 9-9 shows the galvanic cell setup to measure the concentration of
the analyte K+in the solution. The indicator electrode consists of a reference
Ag/AgCl electrode in a solution with a known KC1 concentration (e.g., 0.1
M KCl). A calomel electrode is used as the other reference electrode.
The cell diagram of this cell can be denoted as:
Hg(1) 1 Hg~Ch(s)( C1-(sat'd) 1) K+analyte ( membrane 1 K'(O.l M), C1 I AgCl(s) I Ag(s)
The cell galvanic potential of this setup is:
+
Ecell = Ereference + Eindicator = Ecalorne~+ E A ~ / A ~ /Emembrane
CI
Since both Ecalomel and E A ~ I Aare
~IC known,
~

where pK = -log[K+].
A plot of the cell potential Eeu versus pK at 25OC is a straight line with a
negative slope of 0.059.

pH Electrode
Another example of an ISE is the glass membrane pH electrode. The
membrane of this electrode is usually a thin (0.1 mm) sodium glass with a
composition of 72% SiOn, 22% NanO, and 6% CaO. The silicon and the oxy-
gen in the glass membrane form a negatively charged structure with mobile
positive ions to balance the charge. When the glass is soaked in an aqueous
solution, the aqueous solution exchanges H' for Na+ at the glass surface. The
amount of Na+-H+ exchanged is proportional to the H+ concentration in the
solution.
When the two sides of the glass membrane are soaked in two solutions
of different H' concentrations, a potential Eglass develops across the glass
membrane. At 25OC,

If one of the solutions has a known constant H+ concentration, the con-

centration of H+in the other solution can be determined.
Figure 9-10 shows a setup to measure the pH of a solution. The cell dia-
gram of this cell can be denoted as:
Hg(1) I HpClz(s) I C1-(sat'd) 11 H+analyteI glass I H'(0.1 M), C1- ( AgCl(s) ( Ag(s)
 Electrochemical Transducers

Similar to pK electrode, the cell potential is:

Eceu = C' + 0.059 log [H+analyte]= C' -0.059pH
where flH = -log[H+].
A plot of the cell potential Eccll versus flH at 25OC is a straight line with a
negative slope of 0.059.
In a base solution, sodium glass starts to react with alkali metals such as
in ad* to Hh.A 5x1-thmnglass m-embnne $~-ectmde will s't;lxt 'to pro-
duce noticeable error (referred to as alkaline error) at pH above 9. Therefore,
to measure solution with pH greater than 9, a lithium glass membrane (72010

Calomel
reference
electrode

Figure 9-10. Measurement of pH Using a Glass Membrane Electrode.

SiOz, 22% LizO, and 6% CaO) is used instead.

The reference electrode and the H+sensing glass membrane electrode in
most industry models are combined into a single probe. Ag/AgC1 electrodes
may also be used (to replace the calomel electrode) as the reference elec-
trode. A voltmeter connected to the electrode is usually calibrated to provide
a direct readout of the measured pH. As pH is affected by temperature, pH
meter is often calibrated by using two known buffers with known tempera-
ture characteristics.

pCOz Electrodes
In pCOs measurement, a membrane permeable to COz (e.g., silicon rub-
 Biomedical Device Technology: Pfinct$les and Design

ber) is used to separate the sample solution (e.g., blood) from a buffer (e.g.,
sodium bicarbonate). As the COa diffuses from the sample into the buffer,
the pH of the buffer is lowered. The change in pH in the buffer, measured
by a pH electrode, correlates to the pC0z in the sample. A voltmeter is often
connected across the pH electrode and a reference electrode (Figure 9-1 1) to

Calomel pH
electrode

C02 permeable
membrane

Blood sample inlet

Blood sample in /
contact with membrane
Figure 9-1 1. Measurement of pCOs in Blood.

produce a reading calibrated to read pCOz.

pOz Electrodes
In a pOz electrode, a constant voltage source is applied across and the
current flowing through the electrode is measured. The magnitude of the
current is a function of the pOz level in the sample solution. This electro-
chemical method is known as amperometry (measurement of current flow-
ing through the electrodes with an applied voltage) in contrast to potentiom-
etry (measurement of voltage between the electrodes with almost no current
flow). An amperometric cell is also known as a polarographic cell.
Figure 9-12 shows a Clark polarographic oxygen electrode designed to
measure pOz in a solution. The indicator electrode is a platinum cathode in
a buffer solution of KC1. The anode is a Ag/AgCl electrode. An oxygen-per-
meable membrane (such as Teflonm or polypropylene) separates the buffer
and the sample solution. When a voltage is applied across the anode and the
cathode, oxygen molecules diffused through the membrane are reduced at
the platinum cathode to form OH- ions. The half-reactions at the electrodes
 Electrochemical Transducers

are:
Ag(s) + C1-(aq) + AgCl(s) + e- Anode
Cathode
The magnitude of the current flowing out from the cathode is proportional
to the dissolved Oz level in the sample solution.
The cathode of a p02 electrode is usually made from an inert metal such
as gold or platinum. To increase the respond time, oxygen must diffuse rapid-
ly through the membrane and reach the cathode quickly. To achieve this, the
membrane is made to be very thin (about 20 pm) and the cathode is in the
form of a disk placed very close to the membrane (with separation about 10

Blood
sample inlet

P-4-
- --
Blood sample in /
contact with membrane
Platinum disk (cathode) in
close proximity to membrane

Figure 9-12. Measurement of pOa in Blood Serum.

A galvanic oxygen cell operates on the same principle, except that it does
not have an external voltage source. Instead of measuring its current, the cell
voltage is monitored. The cell potential is directly proportional to the con-
centration of oxygen of the gas outside the membrane.
pH, pCOz, and pOz are the three analytes in blood gas analysis. In prac-
tice, the membranes of these electrodes must be kept clean and replaced reg-
ularly. Two point calibrations of the electrodes should be performed at 37OC
and corrected to 37OC saturated vapor pressure.
 Biomedical Device Technology: Principles and Des@

BIOSENSORS (ENZYME SENSORS)

An oxygen sensor (e.g., Clark oxygen electrode) can be used indirectly

to measure glucose or cholesterol concentration in blood. In serum glucose
measurements, glucose in blood serum is first broken down by the enzyme
glucose oxidase. This process consumes oxygen (see reactions below). The
concentration of glucose is proportional to the amount of 0 2 consumed,
which can be measured using an oxygen electrode. The reaction is:
Glucose + Oz + H z 0 + gluconic acid + H202
In cholesterol measurement, the cholesterol ester in the blood serum is
first broken down into cholesterol by the enzyme cholesterol esterase and
then into cholest-4-ene-3-one and H202 by the enzyme cholesterol oxidase.
The latter reaction consumes oxygen. The amount of 0 2 consumed is pro-
portional to the concentration of the blood serum cholesterol.
Cholesterol esters +Ha0 + cholesterol + fatty acid
Cholesterol + 0 2 + cholest-4-ene-3-one+ Hz02
In addition to measuring the oxygen consumed, an alternative method is
to measure the amount of hydrogen peroxide produced from the reaction,
which is proportional to the concentration of glucose or cholesterol in the
blood serum. The hydrogen peroxide is separated by an ion selective mem-
brane (such as cellulose acetate membrane) and oxidized to give oxygen
according to the reaction:

The amount of 0 2 , which is proportional to the concentration of glucose, is

measured by an oxygen sensor.
In general, an enzyme sensor consists of three layers:
1. The first (outer) layer allows the analyte (e.g., glucose) to pass from the
sample solution (e.g., blood serum) into the electrode.
2. The second layer contains an enzyme (e.g., glucose oxidase) to pro-
duce H202.
3. The inner layer collects and oxidizes Hz02 to form 0 2 . An Oz sensor
in this layer generates a current proportional to the concentration of
the analyte.
Other analytes may be measured using this type of biosensor as long as an
enzyme can be found that specifically oxidizes the analyte to produce hydro-
gen peroxide.
 Electrochemical Transducers

BATTERIES AND FUEL CELLS

Some galvanic cells are used in the industry and home as power sources.
These energy storage cells are often referred to as batteries. A common bat-
tery is the Leclanchk cell, also known as dry cell, or zinc-carbon cell. This
galvanic cell consists of a zinc can, which serves as the anode; a central car-
bon rod is the cathode. The anode and the cathode are separated by a paste
of manganese oxide, carbon, ammonium chloride, and zinc chloride moist-
ened with water. At the anode, zinc is oxidized and at the cathode, MnOz is
reduced. The half-reactions are:
Zn(s) Zn" + 2e- Anode
e- + NH4++ MnOz(s) + MnO(0H)(s) + NH3 Cathode
An external conductor connecting the anode and the cathode allows the flow
of electrons from the anode to the cathode through an external load.
A lead acid cell is another example of an energy storage galvanic cell.
The anode of the cell consists of a frame of lead filled with some spongelike
lead. When the lead is oxidized, Pb2+ions immediately precipitate as PbS04
and deposit on the lead frame. The cathode is also a lead frame filled with
PbOz. The half-reactions are:
Anode
2e- + PbOa(s) + 2H' + H S O r + PbS04(s) + 2H20 Cathode
When an external conductor is connected between the anode and the cath-
ode, a current is drawn. The solid PbS04 is produced at both electrodes as
the cell discharges; H+and HS04- are removed from the solution at the same
time.
A lead acid cell can be recharged by imposing a slightly larger reverse
voltage on the cell. This reverse voltage forces the electrons to flow into the
anode and out of the cathode to reverse the reactions, converting PbS04
back into Pb at the anode and into PbOz at the cathode. Cells that are specif-
ically designed for reuse are called secondary cells, whereas those for single
use are called primary cells. Some common primary and secondary cells and
their nominal open circuit voltages are listed in Table 9-2.
A fuel cell is a galvanic cell in which the reactants are continuously fed
into the cell to produce electricity. Figure 9-13 shows one type of hydrogen-
oxygen fuel cell. The electrodes are made of porous carbon impregnated
with platinum (as a catalyst).Hydrogen is oxidized at the anode and oxygen
is reduced at the cathode. The half-reactions are:
Anode
 Biomedical Device Technology: Princ$les and Design

Table 9-2.
Common Primary and Secondary Cells
W e Nominal cell voltage
Primary Cells Zinc-carbon
Zinc-air
Alkaline
Lithium-manganese dioxide
Secondary Cells Lead-acid
Nickel-cadmium
Nickel metal-hydride
Lithium-ion
- - - - -

+ +
4 e Oz(g) 2H20 + 4 0 H - Cathode
Hydrogen and oxygen are combined to produce water (steam). The
process produces an electron flow when an external load is connected
between the anode and the cathode. The overall reaction is:

For the hydrogen-oxygen fuel cell, if the water is present as a liquid, then
E" = 1.23 V; if it is a gas, then E0= 1.18 V. This is the voltage of a single cell.
Practical fuel cells are built from a number of single cells in series or parallel

Anode 7 if Cahtode

Figure 9-13. Hydrogen-Oxygen Fuel Cell.

 El'ectrochemical Transducers

to provide the desired voltage and power output.

There are many types of fuel cells using different electrolytes (e.g., alka-
line, solid polymer, phosphoric acid) and different fuels (e.g., methane, car-
bon monoxide) "burned" at the anode. However, oxygen from air is the oxi-
dant at the cathode in most of these cells. Fuel cells are considered as the
energy conversion of choice in the future as it has no objectionable by-prod-
ucts and relatively high efficiencies (above 80% theoretical efficiency).
 Chapter 10

BIOPOTENTIAL ELECTRODES

OBJECTIVES

State the mechanism of ion flow in the body.

Define biopotential and its origin.
Explain the formation of cell membrane potential and action potential.
List the characteristics of ideal biopotential electrodes.
Explain half-cell potentials, offset potential, and their significance in
biopotential measurements.
Analyze the characteristics of perfectly polarized and perfectly nonpo-
larized electrodes.
Sketch and analyze the electrical equivalent circuit of Ag/AgCl skin
electrodes.

CHAPTER CONTENTS

1. Introduction
2. Origin of Biopotentials
3. Biopotential Electrodes

INTRODUCTION

An Italian physiologist and physicist, Luigi Galvani, was the first to

explore electrical potentials of the body, now commonly called biopoten-
tials. The element that produces electrical events in biological tissue is the
ion in the electrolyte solution, as opposed to the electron in the electrical cir-
 Biopotential Electrodes

cuit. A biopotential, then, is an electrical voltage caused by a current flow of

ions through biological tissues. The devices that pick up these biopotentials
are referred to as biopotential electrodes. Biopotential electrodes are a form
of transducer.

ORIGIN OF BIOPOTENTIALS

There are two fundamental mechanisms of ion flow in the body: diffu-
sion and drift. Fick's law states that if there is a high concentration of parti-
cles in one region and they are free to move, they will flow in a direction that
equalizes the concentration. The resulting movement of these charges is
called diffusion.
The movement of charged particles (such as ions) that is due to the force
of an electric field (the forces of attraction and repulsion) constitutes particle
drift. Each cell in the body has a potential voltage across its membrane and
the cell content, known as the single-cell membrane potential Vm. The mem-
brane potential forms the basis for the biopotentials of the body. Some of the
biopotentials of interest include the electrocardiogram (ECG), electroen-
cephalogram (EEG), electrooculogram (EOG), electroretinogram (ERG),
and electromyogram (EMG).
The potential is the result of the diffusion and drift of ions across the
high-resistance semipermeable cell membrane. The ions are predominantly
sodium [Na'] ions moving into the cell, and potassium [K+] ions moving out
of it (Figure 10-1). Because of the semipermeable nature of the membrane,
Na' ions are partially restricted from passing into the cell. As a result, the
concentration of Na+ outside the cell is higher than that inside.
In addition, a process called sodium-potassium pump keeps sodium
largely outside the cell and potassium ions inside. In the process, potassium
is pumped into the cell while sodium is pumped out. The rate of sodium
pumping out of the cell is about two to five times that of potassium pumping
into cells. In the presence of the offsetting effects of diffusion and drift and
the sodium-potassium pump, the equilibrium concentration point is estab-
lished when the net flow of ions is zero. As there are more positive ions
moved outside the cells (Na+)than positive ions moved into the cells (K'),the
inside of the cell is less positive than the outside and more negative ions are
present within the cell. Therefore, the cell is negative with respect to the out-
side; the cell becomes polarized. This potential difference between the inside
and outside of the cell at equilibrium is called the resting potential. The mag-
nitude of the resting potential is -70 mV to -90 mV.
If, for any reason, the potential across the cell membrane is raised, say,
by voluntary or involuntary muscle contractions, to a level above a stimulus
 Biomedical Deuice Technology: Principles and Design

Na-K pump
-4 Diffusion

Cell

Figure 10-1. Mechanism of Cell Resting.

threshold, the cell membrane resistance changes. Under this condition, the
nature of the cell membrane changes and becomes permeable to sodium
ions. The sodium ions will start to rush into the cell. The inrush of positive-
ly charged sodium ions caused by this change in cell membrane resistance
gives rise to a change in ion concentrations within and without the cell. The
result is a change in the membrane potential called the action potential
(Figure 10-2).
During this time, the potential inside the cell is 20 to 40 mV more posi-
tive than the potential outside. The action potential lasts for about 1 to 2 mil-
liseconds. As long as the action potential exists, the cell is said to be depo-
larized. Under certain conditions, this action potential disturbance is propa-
gated from one cell to the next, causing the entire tissue to become depolar-
ized. Eventually the cell equilibrium returns to its normal state (i.e., to its
polarized state) and the -90 mV cell membrane potential is resumed. The
time period when the cell is changing its polarization is called the refractory
period. During this time, the cell is not responsive to any stimulation.
When cells are stimulated, they generate a small action potential. If a
large group of cells is stimulated simultaneously, the resultant action poten-
tials can be readily detected. For example, when the heart contracts and
relaxes, the polarization and repolarization of the heart cells create a resul-
tant action potential. This action potential can be monitored by external
machines using electrodes placed on the surface of the body. This sequence
of polarization and repolarization gives rise to a complete waveform known
as the electrocardiogram.
 Biopotential Electrodes

A
Action potential

b
Time (ms)

Resting potential

-70 mV

Figure 10-2. The Action Potential.

BIOPOTENTItZL ELECTRODES

Biopotential electrodes are transducers that pick up the body electrical

signals via the body electrolytes. An electrode provides the interface that is
needed between the ionic current flowing in the body and the electron cur-
rent flowing in the machine. In other words, the electrode allows the
machine to measure the electrical effects in tissue by making a transforma-
tion between ionic conduction to electron conduction at the tissue-electrode
interface. Ideally, the amount of current flowing through the electrode is
zero; hence the purpose of the electrode is to measure the potential at the tis-
sue due to the flow of ions across the biological tissue. But practically speak-
ing, some small amount of current will flow through the electrode.
An ideal electrode is characterized by a number of features, including:
absence of distortion or electrical noise
immunity to external interference
inertness of the electrode in the presence of tissue and bodily fluid
absence of interference with or influence on the tissue
absence of interference with or influence on the movement of the
subject
ease of making contact with the biological source
invariance of the contact even during long periods of time
absence of discomfort to the subject
repeatability of results
low cost, durability, small size, and low weight
 Biomedical Device Technology: Principles and Design

The Tissue-Electrode Interface

Given the listed preferred characteristics, biopotential measurement
using a pair of electrodes should not be considered as simply connecting two
pieces of wire to two conductors and placing them onto a patient.
Biopotential measurement at the tissue-electrode interface in the clinical set-
ting is a rather complicated process. The electrode is the metal that makes
contact with the electrolyte, and the electrolyte is interfaced to the biological
tissue through which ions are free to flow. The tissue-electrode interface has
a significant effect on the quality of the biopotential signal measured by the
medical device.

Electrode Half-Cell Potential

In Chapter 9, it was discussed that if two dissimilar metals are submerged
into a solution of electrolyte, a potential difference can be measured between
the two metals. Each of these metals with the electrolyte creates a half-cell
potential. In biopotential measurements, each electrode-tissue interface
therefore will create a half-cell potential.
In electrophysiology studies (study of the electrical activities within the
human body) the difference in half-cell potentials that can be detected
between two electrodes is referred to as the "offset potential" of the electrode
pair. Under ideal conditions, when measuring biopotential using a pair of
identical electrodes, the offset potential is zero as the half-cell electrode
potential at each electrode-tissue interface will cancel out each other.
However, in practice, the offset potential is never zero as the half-cell poten-
tials at different locations are never identical. Offset potential on the order of
+0.1 V is very common in body surface biopotential measurements.

Polarized and Nonpolarized Electrodes

As with any redox reaction, at any electrode/electrolyte interface, the
electrode tends to discharge ions into the solution, and the ions in the elec-
trolyte tend to combine with the electrode. That is,
Metal + electrons + metal ions (oxidation reaction)
Electrons + metal ions + metal (reduction reaction)
The net result of these reactions at the tissue-electrode interface is the
creation of a charge gradient, the spatial arrangement of which is called the
electrode double layer. Electrodes in which no net transfer occurs across the
metal-electrolyte interface are said to be perfectly polarized or perfectly non-
reversible electrodes; that is, only one of the two chemical reactions
 Biopotential Electrodes

described above can occur. An example of this is shown in Figure 10-3A.

Electrodes in which unhindered transfer of charge is possible are said to be
perfectly nonpolarized or perfectly reversible; that is, both of the equations
referred to above can occur with equal ease. An example of this is shown in
Figure 10-3B. Perfectly nonpolarized electrodes are the ideal electrode of use
since they allow for the best ion-electron interface. However, practical elec-
trodes used in clinical situations have properties that lie between these ideal
limits.

(A) METALLIC ELECTRODE

Perfectly Polarized Electrodes: Only one reaction occurs with ease.

ELECTRODE DOUBLE LAYER FORMS

Potential exists between the electrode and electrolyte
due to the formation of the electrode double layer.

(B) SILVEWSILVER CHLORIDE ELECTRODE

Perfectly Nonpolarired Electrodes: Both reactions occur with ease.

Silver chloride forms free silver ions (Ag+) and chloride ions
(CI-) which prevent the formation of the electron double layer
Figure 10-3. Perfectly Polarized and Perfectly Nonpolarized Electrode.
 Biomedical Device Technology: Principles and Design

Silver/ Silver-Chloride Electrodes

The silvedsilver-chloride (Ag/AgCl) electrode is an important electrode
in biopotential measurement since its interface characteristic is close to a per-
fectly nonpolarized electrode. Silver/silver-chloride electrodes may be man-
ufactured in one of two ways: the electrode may consist of a solid silver sur-
face coated with a thin layer of solid silver-chloride, or it may consist of sil-
ver powder and silver-chloride powder compressed into a solid pellet. In
either case, the presence of the silver-chloride allows the electrode to behave
as a near-perfect nonpolarizable or reversible electrode.
When the electrode is submerged in an electrolyte such as potassium
chloride (KCl) containing chloride ions (Cl-), dissociation of silver ions (Ag")
from the electrode into the solution and association of the chloride ions (C1-)
from the solution to the electrode will form free ionic and charge movement
between the electrolyte and the electrode. This free ionic flow effect pro-
hibits the formation of the electrode double layer. The net result is a low-
impedance, low-offset potential interface between the silver and the elec-
trolyte. Another advantage of the Ag/AgCI electrode is, since silver-chloride
is almost insoluble in a chloride-containing solution (i.e., the body that con-
tains chloride ions), very few free silver ions exist. Therefore, tissue damage
as a result of the silver ions is negligible.

Electrical Equivalent Circuit

One of the most common electrodes used in biopotential measurement
is the surface (or skin) electrode. A typical example of surface electrode is the
electrode used in electrocardiogram acquisition. In such application, the
electrodes are placed on the surface of the skin over the chest; each pair of
electrodes is placed at some specified location and connected via lead wires
to the input of a differential amplifier (Figure 10-5). The difference in poten-
tial at the two electrode sites is amplified and displayed.
An approximate, electrical equivalent circuit of a tissue-electrode inter-
face is shown in Figure 10-4. In this model, the impedance to ion flow
between the tissue and the electrode through the skin is modeled as a 1,000
0 resistor, often referred to as the skin resistance ( R e ) . The double-layer
impedance may be regarded as a 10 yF capacitor (Cd) in parallel with a rea-
sonably high value resistor (Rd), typically 10 kll. V h c represents the half cell
electrode potential; about 0.3 V for a typical Ag/AgCl electrode and KC1
electrolyte.
Figure 10-5 shows the configuration and its electrical equivalent circuit
of a pair of electrodes attached to the patient's body and connected to a dif-
ferential amplifier. The upper and lower branches represent the two elec-
 Biopotential Electrodes

Cd
Figure 10-4. Electrical Equivalent Circuit of Single Electrode Tissue Interface.

trode-tissue interfaces; a 2 MLR resistance represents the input impedance of

the differential amplifier; the impedance to ion flow in the tissue between the
electrodes is modeled by a 100 LR resistance.

II Half cell

10 kOhm
2 MOhm

Tissue Skin
To amplifier impedance resistance input
impedance
II
Double layer
impedance

Iconic current Electron current

Figure 10-5. Electrical Equivalent Circuit of Biopotential Measurement.

Note that in practice, the half-cell potential of the two electrodes will not
be identical; therefore, a nonzero DC offset voltage will appear at the input
of the differential amplifier and hence will appear as an amplified DC offset
voltage at the output of the amplifier. In some applications, this DC offset
can be eliminated by using a high pass filter. The value of the double layer
capacitance (Cd) is larger for a polarized electrode and smaller for a nonpo-
larized electrode (as there is little static charge accumulated at the electrode-
electrolyte interface).
 Biomedical Device Technology: Princ@les and Design

Floating Surface Electrode

At the electrode-tissue interface, any mechanical disturbances will create
electrode noise. This is especially true of surface electrodes because the elec-
trode double layer acts as a region of charge gradient. Figure 10-6A shows a
Ag/AgCl electrode resting directly on the skin surface with a thin layer of
electrode gel. Electrode gel is a jelly pastelike solution containing chloride
ions; its primary function is to provide a good electrical conduction interface
between the electrode and the tissue. Any relative movement between the
electrode and skin surface will create a disturbance to the charge gradient
distribution in the gel. Any disturbance on this charge gradient causes a
change in capacitance and thus a change in the measured electric potential.
The result of this disturbance is referred to as motion artifacts in electro-
physiological measurements.
The electrical stability of the electrode may be considerably enhanced by
mechanically stabilizing the electrode-electrolyte interface. This is achieved
by using indirect-contact floating electrodes that interpose an electrolyte jelly
paste or gel between the electrode and the tissue. This gel substantially
reduces any electrical noise arising from mechanical disturbances in the dou-
ble-layer charge gradient. Figure 10-6B illustrates the stabilization of the
double layer by using a gel-filled electrode housing. In this configuration,
any movement between the electrode housing and the skin surface will cause
little or no disturbance to the electrode double layer as it is at a distance away
from the location of movement. A gel-filled foam between the electrode and
tissue will have a similar effect. In the case of internal electrodes, which are
usually constructed of stainless steel, the electrode-tissue interface is already
enhanced and stabilized by the presence of extracellular fluid.

Gel housing -----.

.---------
AgIAgCI electrode - - - - - - - -
1 .*
,----.--
Electrode gel - -- - - -
/j-q-&, ----
k.
. Skin surface -------0'

(a) Direct contact (b) Gel-filled

Figure 10-6. Gel-filledElectrode Housing to Minimize Motion Artifact.
 Biopotential Electrodes

Other Biopotential Electrodes

The Ag/AgCl surface electrode described above is only one of the many
different types of electrodes used in biopotential signal measurements. For
example, metal (e.g., copper) electrodes can be used in measuring surface
ECG. Other than surface electrodes, invasive electrodes are commonly used
to measure signals deep inside the body or in a small localized region in the
tissue. Figure 10-7A shows a needle electrode for measuring electromyo-
gram. It measures the localized electrical activity when inserted into a mus-
cle fiber. Figure 10-7B is a fine-wire electrode for similar applications but can
allow more movement by the subject as it is more flexible than a needle elec-
trode.

R Insulated coating

Skin

Muscle

Uninsulated barb

(a) Needle electrode (a) Fine-wire electrode

Figure 10-7. Invasive Biopotential Electrodes.
 Chapter 11

BIOPOTENTIAL AMPLIFIERS

OBJECTIVES

State the characteristics of an instrumentation amplifier.

Define common mode and differential mode input.
Derive the differential gain, common mode gain, and common mode
rejection ratio of an instrumentation amplifier.
Explain the use of a differential amplifier in minimizing common mode
signals.
List common sources of noise and interference affecting biopotential
measurements in health care environment.
Analyze methods to reduce noise and interference including lead
shielding, input guarding, right leg driven circuits, and filters.
Explain interference due to magnetic induction and describe means to
reduce such interference.
List sources of conductive interference and explain the principles of
surge suppressors.

CHAPTER CONTENTS

1. Introduction
2. Instrumentation Amplifiers
3. Differential and Common Mode Signals
4. Noise in Biopotential Signal Measurements
5. Interference from External Electrical Field
6. Interference from External Magnetic Field
7. Conductive Interference
 Biopotential ArnpliJiers

INTRODUCTION

A biopotential signal is often small in amplitude, mixed with other sig-

nals, and subjected to external interference. In addition to amplifying the sig-
nal, a biopotential amplifier is designed to extract the desired signal from
interfering sources as well as to prevent external noise to corrupt the signal.
This chapter studies the principles and design of biopotential amplifiers to
achieve these objectives.

INSTRUMENTATION AMPLIFIER

Figure 11-1 shows a simple operational amplifier equivalent circuit with

gain = A, input voltage K = I
4- K, output Ilo = AK, input impedance Zi, and
output impedance Z .

Figure 11-1. Operational Amplifier.

An ideal operational amplifier should have the following characteristics:

Infinite open loop gain, or A = a
Infinite input impedance, or Zi = 03
Zero output impedance Zo = 0
Infinite bandwidth and no phase distortion
In reality, there is no ideal operational amplifier. However, in circuit
analysis, when compared to other circuit parameters, a good operational
amplifier may be considered to be ideal, which, to a great extend, simplifies
the analysis process.
In biopotential measurements, a special amplifier called the instrumen-
tation amplifier is often used. An instrumentation amplifier (Figure 11-2) is
a closed-loop, differential input amplifier designed with the purpose of accu-
rately amplifying the voltage applied to its inputs. Ideally, an instrumentation
 Biomedical Device Technology: Principles and Design

Figure 11-2. Instrumentation Amplifier.

amplifier responds only to the difference between the two input signals and
has very high input impedance between its two input terminals and between
each input to ground. The characteristics of a good instrumentation amplifi-
er are:
Very large input impedance
Very low output impedance
Constant differential gain with zero nonlinearity
High common mode rejection
Very wide bandwidth with no phase distortion
Low DC offset voltage or drift
Low input bias current and offset current
Low noise
Figure 11-3 shows a single Op-Amp differential amplifier.

-
4
Figure 11-3. Differential Amp Stage of Instrumentation Amplifier.

The voltage V+at the noninverting input terminal of the amplifier is

given by:
 Biopotential Amplijiers

For an Op-Amp in its active nonsaturated state, the voltages at the input
terminals of the amplifier must be the same, i.e., V+= V-. The current flow-
ing through RI and R3 are given by:

Since the input impedance of the Op-Amp is very large, there is no cur-
rent flowing into the input of the amplifier; the currents flowing through RI
and R3 are the same, therefore:

Solving equations (I) and (2) using V- = V+yields:

v, = [RP W]v4- [BL] v3
Ri (Fb + R4)
R1
If -= Rz
-, equation (3) can be simplified to:
R3 R4

Equation (4) is the characteristic of a true differential amplifier with dif-

ferential gain DG equal to -R3/Rr and common mode gain CMG (when Vj
= TJ4) equal to zero. The common mode rejection ratio (CMRR) of this
amplifier is therefore equal to infinity:

C M R R = ~ -D~G ~
=-=m oR I
CMG
The input impedance of this amplifier between the inverting and nonin-
verting input to ground is RI+ R3 and Rz + R$ respectively. Since these resis-
tances must be much smaller than the input impedance of the Op-Amp, typ-
ical input impedance of this differential amplifier stage is usually chosen to
be below 100 kfl. To overcome this shortcoming, another amplifier stage
shown in Figure 11-4 is required to increase the input impedance of the
instrumentation amplifier. Below is the analysis of this impedance matching
stage of the amplifier:
 Biomedical Device Technology: P/inc$les and Design

v4
v2
Figure 11-4. Input Stage of Instrumentation Amplifier.

For an ideal Op-Amp, due to its large input impedance, no current flows
into the input terminals; therefore, the currents flowing through 125, Rs, and
R7 are identical and can be written as:

Since the voltages at the two input terminals of an active nonsaturated

Op-Amp are identical, the current flowing through the resistor Rs is equal to:

Therefore,

v3 - v4 = R5 + + R7 (V, - Vn)
Rc,

The differential gain of this stage is therefore =

R5 + R(i + R7 . From
R(i
equation (6),when F5 = Vz, i = 0. Since there is no current flowing through
the R5, Ilj = VI;similarly, we can show that Vb = Vz. If the common mode
voltage at the input is Vn, the same common voltage will appear at the out-
puts Vz and fi of the Op-Amps. Therefore, the common mode gain for this
circuit is equal to unity. The common mode rejection ratio for this stage is
given by:
 Biopotential AmpliJiers

As the signals inputs are directly connected to the input terminals of the
Op-Amp, the input impedance of the circuit is equal to the input impedance
of the Op-Amp. Figure 11-5 shows the combination of these two amplifier
stages forming a classical instrumentation amplifier (IA).The theoretical dif-
ferential gain, common mode gain, common mode rejection ratio, and input
impedance of this classical IA are:

CMG = zero,
CMRR = infinity, and
Zi = infinity.
However, these values are not achievable due to nonideal Op-Amp char-
acteristics (such as nonzero input bias and offset current). In practice, a good
IA can have CMRR > 100,000 and Zin > 100 M a .

Figure 11-5. Classical Instrumentation Amplifier.

DIFFERENTIAL AND COMMON MODE SIGNALS

Consider the instrumentation amplifier shown in Figure 11-6a with volt-

age signals VIand V2 connected to the input terminals of the amplifier. These
input signals can be represented by their common mode and differential
mode signals, I4 and V4 respectively (Figure 11-6b). The mathematical rela-
tionships between these voltages are shown by the following equations:
 Biomedical Device Technology: Princ$les and Design

Figure 11-6. Common Mode and Differential Mode Inputs.

If the amplifier is an ideal instrumentation amplifier with a differential

gain of A, the output is equal to AVL The common mode signal K will not
appear at the output (since the common mode gain is zero).

Example 7
Referring to the amplifier in Figure 11-2, suppose the voltage measured at
V I with respect to ground is 4.0 mVdc, the voltage a Vz with respect to
ground is 2.5 mVdc, and if the differential gain Ad of the differential ampli-
fier is 500, the output voltage Vout of the differential amplifier is:

Solution
Vout= Ad (Va - VI) + 500 (4.0 - 2.5) mVdc = 500 X 1.5 mVdc = 750 mVdc
 Biopotential AmpliJiers

Example 2
For an instrumentation amplifier with differential gain Ad = 1,000, common
mode gain Ac = 0.001, what is output voltage Voutif the differential input is
a 1.5 mV 1 Hz sinusoidal signal and the common mode input is 2.0 mV 60
Hz noise?

Solution
Vd = 1.5 sin (2nt) mV, Vc = 2.0 sin (120nt) mV
+
VoUt= AdVd + AcVc= [1,000 x 1.5 sin (2nt) 0.001 X 2.0 sin (12Ont)lmV
= [1,500 sin (2nt) + 0.002 sin (120nt)l mV
= 1.5 sin (2nt) V + 2.0 sin (120nt) pV
This example illustrates the function of the differential amplifier to amplify
the differential (desired) while suppressing the common mode signal (noise).

In an experiment with the Op-Amp in Figure 11-6b, if Vout= 10 V when Vd

= 1.0 mV and Vc = 0.0, and Vout = 50 mV when Vd = 0.0 and Vc = 5.0 V,
find the differential gain A, the common mode gain Ac, the common code
rejection ratio CMRR, and CMRdB.

Solution
Ad = VouJVd = 10/0.001 = 10,000, Ac = VouJVc = 0.05/5 = 0.01
CMRR = Ad/Ac = 10,000/0.01 = 1,000,000
The CMRR expressed in dB (that is, CMRdB) is given as:
CMRdB = 20 log (CMRR) = 20 log (1,000,000) = 120dB.

NOISE IN BIOPOTENTIAL SIGNAL MEASUREMENTS

Biopotential signals are produced as a result of action potentials at the

cellular level. In physiological monitoring, a biopotential signal is often the
resultant electrical potentials from the activities of a group of tissues. We have
learned in earlier chapters that a biopotential signal is usually small in mag-
nitude and surrounded by noise. One of the many problems with the ampli-
fication of small signals is the concurrent amplification of noise and the pres-
ence of interference. Interference and noise are simply defined as any signal
 Biomedical Device Technology: Principles and Design

other than the JesireJsignar. rn physibrogicar signar measurements, there are

two different sources of noise and interference:
1. Artificial sources from the surrounding environment such as electro-
magnetic interference or mechanical motion. For example, artifacts
on an ECG strip caused by fluorescent lighting or unshielded power
supply voltages are considered artificial interference.
2. Natural biological signal sources from the patient. For example, in
ECG measurement, any signals other than ECG that arise from other
biopotentials of the body are considered as natural noise. These
include muscle artifact from the patient or electrical activity of the
brain. While brain activity is noise when measuring ECG, ECG sig-
nal is considered as noise when brain wave (EEG) is measured.
One of the functions of a biopotential amplifier and its associated electronic
circuitry is to amplify those biopotential signals of interest while rejecting or
minimizing all other interfering signals.
An example of a common form of interference in biopotential measure-
ments is shown in Figure 11-7. The ECG signal is corrupted by 60 Hz power
noise induced on the body of the patient. Noise from power line interference
may have an amplitude of several millivolts, which can be larger than the sig-
nal of interest. Fortunately, the power line induced 60 Hz noise is on the
entire body of the patient and therefore appears equally at both the invert-
ing and noninverting input terminals of the biopotential amplifier. Such com-
mon mode signal can be substantially reduced by using a good instrumenta-
tion amplifier with a large common mode rejection ratio. Filters can also be
used to remove the undesirable signal if the bandwidth of the interfering sig-
nal is not overlapping that of the desired signal.
In order to obtain a good signal in a noisy environment, it is important
to have a signal level much larger than the noise level. The ratio of signal to
noise level is an important parameter in signal analysis and processing.

Figure 11-7. 60Hz Power L n e Interference on ECG Signal.

 Biopotential AmpliJiers

Signal to noise ratio (SNR) in decibel is defined as:

SNR(dB) = 20 X
vs
log -
vn
where K and Vi are the signal and noise voltage, respectively.
When dealing with medical instrumentations, the most common external
noise source is from the power lines or equipment in the patient care area.
Interference from 60 Hz power can be induced by electric or magnetic fields.
A 60 Hz electric field can induce current on lead wires as well as on the
patient's body. A changing magnetic field (e.g., from 60 Hz power lines) can
induce a voltage or current on a conductor loop. Other than 60 Hz power
line interference, much equipment (e.g., switching regulators, electrosurgical
units) emits electromagnetic noise into the surrounding area. These EM1 can
be of low or high frequencies (e.g., 500 kHz from an electrosurgical unit),
which may create problems if it is not dealt with properly. EM1 can be radi-
ated as well as conducted through cables or conductor connections. For
example, high-frequency harmonics from switching power supplies can be
transmitted through the power grid to other equipment in the vicinity.
Switching transient, which may cause damage to electronic components, can
be transferred in the same way.
In general, the design of the first stage of medical devices, which usually
includes the patient interface and the instrumentation amplifier, is critical to
maintain a healthy SNR. The remainder of this chapter discusses the mech-
anism of interference and some practical noise suppression measures.

INTERFERENCE FROM EXTERNAL ELECTRICAL FIELD

A typical electromedical device measuring biopotential from the patient

has conductor wires connecting the device to a patient. The patient and the
device are usually working in an environment filled with electric field pro-
duced by 120 V power lines and line-powered devices. Figure 11-8 shows
such an arrangement. Using an ECG as an example, terminals A and B are
the input of the instrumentation amplifier and G is the ground connecting
point of the device.
Under typical operating conditions, through capacitive coupling, the
electric field produced by the 120 V power sources will induce current flow-
ing into the lead wires as well as into the patient's body to ground. The fol-
lowing sections examine the effects of such induced currents on the output
and methods to reduce these interferences.
 Biomedical Device Technology: Princ$les and Design

A
Medical
Device
B
G

Figure 11-8. Connection of Medical Device to Patient.

Currents Induced on Lead Wires

Consider that the medical device is an electrocardiograph (ECG) with
skin electrodes attached to the patient as shown in Figure 11-9. Zi, Z2, and
ZG represent the impedances of the skin electrode interfaces and electrodes.
C I and Cz are the coupling capacitors between the 120 V power line and the
lead wires. These capacitance values depend on the length of the conductors
and their distance from the 120 V power sources. Due to these capacitive
couplings, displacement currents (Id1 and I ~ zwill
) flow in the wires to ground.
Similarly, C3 is the coupling capacitor between the power line and the chas-
sis of the device. As the chassis is grounded, the displacement current Id3 will
flow through the chassis to ground.
C b is the coupling capacitor between the power line and the body of the
patient. A displacement current Idb will flow into the body of the patient.
Assuming the input impedances of the instrumentation amplifier (IA) are
very large, the displacement currents Id1 and Id2, which are 60 Hz induced
currents, will flow through the skin-electrode interfaces, into the patient's
body and out through the skin-electrode interface (ZG)to ground. This cur-
rent path will create a voltage across the input terminals A and B of the IA
given by:

(note the cancellation of the common mode voltage I ~ B Zby

G the differential
amplifier).
 --7----"-----
-Z- I
r - - 1 - - - - - -T
I I
--
I 120V a.c.
~b -:--I ~ D Id1 iI
162 i
I
I~Gi
I

T
I
143
I I I
cl-d-- -d--CG -J-- --I-.
I

-1-- -1-- -1-- -1-- C 3

I c 2 I I
I
I
I
I
L
\

A
;r' : r'
'
1
I
I
I
ECG
I I
I \; B
r' G
I
I
I
I
I
I
Id0 I
I
L I

Figure 11-9. Interference from Power Line.

If similar electrodes and lead wires are used and they are placed close
together, one can simplify this expression by making Id1 = Id2 = Id. In this
case:
VA- VB= Id (ZI - 2 2 )

Example 4
In an ECG measurement using the setup in Figure 11-9, if the displacement
current Id due to power line interference is 9 nA and difference in the skin-
electrode impedances of the two limb electrodes are 20 k a , find the 60 Hz
interference voltage across the input terminals of the ECG machine.

Solution
Using the above derived equation:

With a typical ECG signal amplitude of 1 mV, this represents 8% of the

desired signal. This electromagnetic interference creates a 60 Hz signal rid-
ing on the 1 mV amplitude ECG waveform (Figure 11-7).
 Biomedical Device Technology: Principles and Design

Example 5
For Example 4, if the displacement current I d b through the patient's body is
0.2pA, what is the common mode voltage at the input terminals of the ECG
machine given that the skin-electrode impedance ZG at the leg electrode is
50 kR and the body impedance Z b is 500R?

Solution
The common mode voltage Vcrn at the input terminals of the ECG machine
is due to the current flowing through the body impedance and the skin-elec-
trode impedance.

Since I d b is much larger than Id1 and Id:! and ZG is much larger than Zb, we
can write
Vcm = I ~ ~ =
Z 0.2
G pA X 50 kR = 10 mV.
This common mode voltage is 10 times the typical amplitude of an ECG signal!
Fortunately, this 60 Hz common mode signal will not appear at the output
due to the high common mode rejection signal of the instrumentation ampli-
fier.

From equation (7), in order to reduce the interference signal (power line
60 Hz interference in this case), it is desirable to reduce I d 1 and I d 2 or ensure
that the skin-electrode impedances are the same (so that ZI - Z z = 0). The lat-
ter can be achieved by using identical electrodes and ensuring that proper
skin preparation is done before the electrodes are applied. One attempt to
reduce or even eliminate I d 1 and I d 2 is to use shielded lead wires as shown in
Figure 11-10a. When the entire length of the lead wires is surrounded by a
grounded sheath, the coupling capacitors between the power line and each
lead wire are eliminated (i.e., I d = I d 1 = 182 = 0). Therefore, from equation (7),
VA - VB = 0. However, the shield, which is in close proximity to the lead
wires, creates coupling capacitances C s l and C s 2 with each of the wires. From
the equivalent circuit shown in Figure 11-lob, one can show that the voltage
across the input terminals of the ECG machine due to the nonzero common
mode voltage on the patient body (see Example 5 for estimation of body
common mode voltage) is equal to:

VA-VB= ( Zsl
Zl+ZSl
-
zs2
z2+zs2
) vcm

where Zsr and 2 2 are the impedances due to capacitances G r and Cs2, respec-
tively.
 Biopotential Amplijiers

z1
Equation (8)becomes zero only if -= -.Z s l If this condition is not met,
2 2 Zs2
Vcm will appear at the output no matter how good the common mode rejec-
tion ratio of the ECG is.

Figure 1 l-10a. Lead Shielding.

ECG

Figure 1l-lob. Lead Shielding-Equivalent Circuit.

To prevent this, a second shield (called the guarding shield) is placed

between the first shield and the lead wires, and the guarding shield is con-
nected to the patient's body (e.g., the right leg) via an electrode. This setup
 Biomedical Device Technology: Princ$les and Design

is shown in Figure 11-lla and its equivalent circuit is shown in Figure

11-llb. Note the potential of the guarding shield is at the same level as that
of the patient body (i.e., at Vcm). Therefore, Vcm will not show up across the
input terminals of the ECG as there is no current flow around the loops of
Zi-Zs'~-Z~and Z~-Z~Y-ZG in the circuit. (Zs~1
and Zss are the impedances of
the coupling capacitors Cs11 and Cs9, respectively, between the guarding
shield and the lead wires.)
By using the input guarding method, the induced lead current I d and the
common mode voltage Vcmwill not appear across the input terminals of the
instrumentation amplifier of the ECG.

I Guarding Shield

Figure 11-lla Input Guarding.

ECG

Figure 11 -1 1b. Input Guarding-Equivalent Circuit.

 Biopotential Amplijiers

Right-Leg-Driven Circuit
So far, we have been assuming that the ECG input stage is an ideal
instrumentation amplifier (i.e., with infinite input impedance and zero com-
mon mode gain). Under these ideal conditions, all common mode signals at
the input terminals of the IA are rejected. Let's consider a more realistic sit-
uation when the input impedance of the instrumentation amplifier has a
finite value.

- Idea
IA
1
ECG

Figure 11-12. Effect of Common Mode Voltage on Finite Input Impedance.

An ECG machine with a nonideal instrumentation amplifier can be rep-

resented by an ideal IA coupled to finite input impedances Zi to ground at
each of its input terminals (Figure 11-12). The differential input voltage of
this configuration is given by:

If Zin is much greater than Zi and Zz,we can simply the equation to:
V A - V B = 22 - Z1 Vcm
Zin

This differential input voltage (Ki - fi) to the IA due to nonzero com-
mon mode voltage Kmwill be amplified no matter how large the CMRR is
(or how small the CMG). To reduce this voltage, we can either choose an IA
with very large Ztn,use perfectly matching electrodes with good skin prepa-
 Biomedical Device Technology: Principles and Design

ration (i.e., make Zr = 22) or reduce Vm. A practical method using active can-
cellation to reduce V l is the right leg driven circuit shown in Figure 11-13.

Example 6
An instrumentation amplifier with input impedances of 5 McR between each
input terminal to ground is used as the first stage of an ECG machine. If the
difference in the skin-electrode impedance is 20 kcR and the common mode
voltage induced from power lines on the patient's body is 10 mV, calculate
the magnitude of the 60 Hz interference appearing across the input terminals
of the IA.

Solution
Substituting the value into equation (9), the voltage across the input of the
ideal IA is:

Since this is a differential input signal to the IA, it will be amplified and

R2
U1
Rt
-
to right
arm
3

~5
r VAI
~3
uh,
.VO
Ra

to left
arm
- R6 '

Rb
U2 R7
-- Rf

=j
b w&
2 -
toig\
leg
RG
- 3D7i

Figure 11-13. Instrumentation Amplifier with Right-Leg-Driven Circuit.

 Biopotential Amplifiers

appear at the output of the ECG machine. Compared to the typical ampli-
tude of an ECG signal (1 mV), this is a noticeable noise level.
The Op-Amps Ul, U2, and U3 form a classical instrumentation amplifi-
er (see also Figure 11-5). The inputs are connected to the left and right arm
electrodes of the patient. The output voltage K of U3 is the amplified biopo-
tential signal between these two limb electrodes. K is coupled to the next
stage of the ECG machine. U4 with &and Ro forms an inverting amplifier
with input taken from the output of U l and U2. This circuit extracts the com-
mon mode voltage from the patient's body, inverts it, and feeds it back to the
patient via the right leg electrode. It creates an active cancellation effect on
the common mode voltage induced on the patient's body and thereby
reduces the magnitude of Vi to a much smaller value. Figure 11- 14 redraws
the right leg driven circuit to facilitate quantitative analysis of this circuit

Figure 11-14. Equivalent Circuit of Right-Leg-DrivenCircuit.

For the right-leg-driven (RLD) circuit in Figure 11-13, Ra is chosen to be

equal to Rb. We have also proven earlier that the same common mode volt-
age Vi at the patient's body appears at the output of U l and U2. Therefore,
we can represent this part of the circuit by a voltage source Vm and a resis-
tor with resistance equal to the parallel resistance of Ra and Rb (equal to Rd2
as Ra = Rb). At the right leg electrode, the voltage at the electrode is also Km
(electrode connecting to the patient's body), and there is an induced 60 Hz
current Ida flowing from the patient into the ECG machine. This equivalent
circuit is shown in Figure 11-14.
At the noninverting input terminal of U4, due to the large input imped-
ance of the amplifier, there is no current flowing into or out of the Op-Amp.
 Biomedical Device Technology: Princ$les and Design

Therefore Ir + I2 = 0.
But I2 = ( V- O)/&and 1 7 = (Vi - O)/Rn/2, therefore

From the lower branch of the circuit:

Combining equations (10) and (11) gives:

vcm = RG Idb
1 + 2fi/&
From equation (12),we can see that if we want to have a small Vim, we must
make the denominator as large as possible. That is, the ratio of R$Ra should
be very large.

Example 7
For the RLD circuit in Figure 11-13, using the values in Example 5 (i.e., Id6
= 0.2 pA, RG= 50 k a : 1) find the common mode voltage V i on the patient's
body if R f = 5 M a and fi = 25 kR. 2) Using this new V i value, calculate the
magnitude of the 60 Hz interference appearing across the input terminals of
the IA in Example 6.

Solution
1) Substituting values into equation (12) gives:

Vcm = 50 kR 0.2 pA = 125 R X 0.2 p 4 = 25 pV

2X5MR
1+ /25 kR
Using the right-leg-driven circuit, we have reduced the Vi from 10 mV to 25
pV, a 400 times reduction!
2) Substituting values into equation (9), the voltage across the input of the
ideal IA is:

This magnitude of noise is negligible when compared to a 1 mV level of the

ECG signal.
The function of Ro is used to limit the current flowing into the Op-Amp
when there is a large Vm. Ro, usually of resistance equal to several M a , is pri-
 Biopotential Amplijiers

manly for the protection of the electronic circuits during cardiac defibrilla-
tion.
An alternative configuration of the RLD circuit is shown in Figure 11-15.
Interested readers may go through a similar derivation to determine the
common modes signal level using this feedback configuration.

Figure 11-15. Instrumentation Amplifier with RLD (alternative).

INTERFERENCE FROM EXTERNAL MAGNETIC FIELD

Another source of interference is magnetic induction from changing

magnetic fields. This can be from power lines and devices that create mag-
netic fields such as large motors or transformers. If such magnetic field pass-
es through a conductor loop, it will induce a voltage T/; proportional to the
rate of change of the magnetic flux @, that is

where @ is the product of the magnetic field B and the area A of the con-
ductor loop perpendicular to B.
Figure 11-16a shows the magnetic field interference during an ECG
measurement procedure. The conductor loop is formed by the lead wires
and the patient's body. If the magnetic field is generated from the ballast of
 Biomedical Device Technolog: Princ$les and Design

a fluorescent light fitting, a 60 Hz differential signal will appear across the

input terminals of the ECG machine. From equation (13), one can minimize
the magnitude of interference by reducing loop area A. A simple approach is
to place the lead wires closer together to reduce the magnetic induction area.
Another method is to twist the wires together (as shown in Figure 11-16b) so
that the fluxes cancel each other (flux generated in a loop is in opposite
polarity to the flux generated in the adjacent loops). Unwanted magnetic
field can also be shielded but often it is relatively expensive to produce a
practical magnetic shield to protect the device from magnetic field interfer-
ence.
We so far have been using 60 Hz power line signals as examples of elec-

Figure 11-16a. Interference Due to Magnetic Field.

ECG

Figure 11-16b. Interference Due to Magnetic Field.

 Biopotential Amplifers

tromagnetic interference (EMI). Other than power frequency interference,

there are many other sources of interference that radiate EM1 to the sur-
rounding area. In medical settings, electrosurgical units that produce 500
kHz EMI, and radio broadcasting and cellular phones that produce EM1 in
the GHz range are just a few of the many examples.

CONDUCTIVE INTERFERENCE

Other than EMI, which is often considered as radiated noise, interfer-

ence can also be caused by unwanted signals conducted to the device via
lead wires, power cables, et cetera. Some of these conductive interference
sources are discussed in the following sections.

High Frequency and Harmonics

This interference appears as high frequency riding on the 60 Hz power
voltage. It is usually caused by poorly designed switch-mode regulators that
return the high-frequency signal into the power lines. These harmonics and
high frequency affect devices connected to the same power grid. They can
be eliminated by placing power line filters at the input of the power~supply
of a device.

Switching Transients
Switching transients are produced when high voltage or high current is
turned on and off by a switch or a circuit breaker. During the interruption of
a switch or power breaker, arcing occurs across the contact of the switch.
This arcing may generate an overvoltage and a short duration of high-fre-
quency oscillation. Switching transients can damage sensitive electronic
equipment if the device is not properly protected. Switching transient dam-
age can be prevented by using power line filters and surge protectors.

Lightning Surges
When lightning strikes a conductive cable (such as a power line, a tele-
phone cable, or network cable) connected to a device, the high voltage and
high power surge will be transmitted through the power grid into the med-
ical device and cause component damage. Surge protection devices with
adequate power capacities are required to protect electromedical devices
from lightning damage.
 Biomedical Device Technolog: Princ$les and Design

Defibrillator Pulses
Medical devices are designed to be safe to patients and operators. Under
normal operation, patients and users are not subjected to any electrical risk
from the medical device. However, there are times a patient can present elec-
trical risk to a device. An example of such an occurrence is when a patient
may need to be defibrillated while an ECG monitor is still connected to the
patient. In this case, a high voltage pulse of several thousand volts may be
transmitted through the lead wires into the ECG monitors. Special high volt-
age protection circuits (referred to as defibrillation protection circuits) are
built into devices that are subject to such risks. Figure 11-17 shows the defib-
rillator protection components of an ECG machine. In the case of a high
voltage applied to the ECG lead wires, the voltage limiting device will clamp
the voltage, to say 0.7 V, to protect the instrumentation amplifiers and other
electronic components in the machine. The resistance R limits the current
flowing into the voltage limiting device.

Current limiting resistors

ECG

oltage limiting devices

ECG leads
Figure 11- 17. Defibrillator Protection Circuit.

The voltage limiting device can be two parallel diodes arranged as in

Figure 11-18a, two zener diodes in series, or a gas discharge lamp. All these
devices can limit the voltage level at the input of the ECG machine. The
transfer function of the voltage limiting circuit (consisting of the voltage lim-
iting device together with the current limiting resistor) is shown in Figure
11-18b.
 Biopotential Ampl@ers

Diodes Zener Gas Discharge Tube

Figure 11-18a. Examples of Voltage Limiting Devices.

-------

b
Input (V)

---

Figure 11-18b. Characteristics of Voltage Limiting Circuits.

For the defibrillator protection circuit shown in Figure 11-17, since the
ECG signal is at the most a few millivolts, one can use a silicon diode (with
turn-on voltage = 0.7 V) as the voltage limiter. Under normal measurement
conditions, the voltage at the input of the ECG machine will be equal to the
ECG signal (about 1 mV amplitude). During defibrillation, although the volt-
age on the patient's body can be several thousand volts, the voltage at the
input terminals of the ECG machine will be limited to 0.7 V, thereby pro-
tecting the electronic components in the ECG from being damaged by the
high voltage of the defibrillator.
 Chapter 12

ELECTRICAL SAFETY
AND SIGNAL ISOLATION

OBJECTIVES

State the nature and causes of electrical shock hazards from medical
devices.
Explain the physiological and tissue effects of risk current.
Differentiate micro and macro shocks.
Define leakage current and identify its sources.
List user precautions to minimize risk from electrical shock.
Compare grounded and isolated power supply systems.
Analyze the principles and shortfalls of grounded and isolated power
systems in term of electrical safety.
Explain the function of the line isolation transformer in an isolated
power system.
Explain the purpose of signal isolation and identify common isolation
barriers.
Describe other measures to enhance electrical safety.
Evaluate the IEC601-1 leakage measurement device and its applica-
tions.

CHAPTER CONTENTS

1. Introduction
2. Electrical Shock Hazards
3. Macroshock and Microshock
4. Prevention of Electrical Safety Hazards
 Electrical Safety and S e a l Isolation

5. Grounded and Isolated Power Systems

6. Signal Isolation
7. Other Methods to Reduce Electrical Hazard
8. Measurement of Leakage Current

Concerned with the increasing use of medical procedures that penetrate

the skin barrier (e.g., catheterization), a number of studies published in the
early 1970s suggested the occurrences of electrocution of patients from low-
level electrical current passing directly through the heart (microshock).It was
also demonstrated from animal studies that a 60 Hz current with level as low
as 20 p4 directly flowing through the heart can cause ventricular fibrillation.
Although the actual occurrence of death due to microshock in hospitals
has never been documented, the potential of such occurrence is believed to
be present. In the interest of electrical safety and risk reduction, hospitals
have implemented both infrastructure and procedural measures to prevent
such electrical shock hazards. Special considerations were given in designing
medical devices to make them electrically safe. This chapter discusses these
safety measures and device designs to prevent electrical shock to patients.

ELECTRICAL SHOCK HAZARDS

Electricity is a convenient form of energy. The deployment of elec-

tromedical devices in health care has advanced patient care, improved diag-
nosis, and expanded the treatment of diseases. However, improper use of
electricity may lead to electrical shock, fire, or even explosion, which may
lead to patient and staff injuries. The heating and arcing from electricity may
cause patient burn or ignite flammable materials such as drapes, alcohol prep
solutions, or even the body hair of patients. Together with enriched oxygen
content in the surrounding atmosphere, fire and explosion hazards are emi-
nent.
When an electrical current passes through a patient, it creates different
effects on tissue. Tissue effects due to electrical current depend on the fol-
lowing factors:
the magnitude and frequency of the current
the current path
the length of time that the current flows through the body
the overall physical condition of the person
 Biomedical Device Technology: Principles and Design

Skin is a natural defense against electrical shock because the outermost

layer of skin (the epithelium) has very low conductivity. The skin is a good
insulator (relatively high resistance) surrounding the more susceptible inter-
nal organs. The resistance of 1 cm2of skin is about 15 kR to 1 MR (note that
the resistance decreases with increasing contact area). However, skin con-
ductivity can increase 100 to 1,000 times (e-g.,become 150 R) when it is wet.
Due to the nature of medical evaluations and treatments, patients are
more susceptible to electrical shocks. Some reasons of the increased electri-
cal hazards are listed below:
In hospitals, skin resistance is often bypassed by conductive objects
such as hypodermic needles, and fluid-filled catheters introduced
through the skin.
ECG/EEG electrodes use electrolyte gel to reduce skin resistance.
Fluid inside the body is a good conductor to current flow.
Patients are often in a compromised situation (e.g., under anesthesia).
They may not be sensitive to or able to react to heat, pain, or other
discomfort.
Clinicians do not necessary have knowledge of electricity and main-
taining a safe electrical environment.
As an example, a patient in an intensive care unit may have several fluid-
filled catheters connected to his or her heart to allow pressure measurements
in the heart chambers. This same patient could be in an electrical bed; be
connected with ECG electrodes, temperature probes, respiration sensors,
and intravenous lines; be covered by a hypothermic blanket; and be con-
nected to a ventilator. All of these connections have the potential to conduct
a hazardous current to the patient.
An electrical current can produce reversible or irreversible damage to
tissue as well as stimulate muscle and nerve conduction. Table 12-1 shows

Table 12-1.
Potential Hazards from Electrical Current (External Contact).
- -

Current Level -
physiological and Tissue Ejict -- -

1 mA Threshold of perception. The person begins to sense the presence of the

current.
5 mA Maximum accepted safe current.
10 mA Maximum current before involuntary muscle contraction. May cause the
person's finger to clamp onto the current source. ("let gon current).
50 mA Perception of pain. Possible fainting, exhaustion, mechanical injury.
100-300 mA Possible ventricular fibrillation.
6A Sustained myocardial contraction; temporary respiratory paralysis; may
sustain tissue burns.
2 10 A All of the above plus severe thermal burns.
- - - - - - - - -
 Electrical Safety and Signal Isolation

the human physiological and tissue effects of a one-second external contact

with a 60 Hz electrical current.

Example 7
A patient is touching a medical device with one hand and grabbing the
handrail of a grounded bed. If the ground wire of the medical device is bro-
ken and there is a fault in the medical device that shorted the chassis of the
device to the live power conductor (120V), what is the risk current passing
through the patient? Assume that each skin contact has a resistance of 25 kR
and the internal body resistance is 500 0.

Solution
If the ground wire of the medical device is intact, a large fault current will
flow to ground and blow the fuse of the device or trip the circuit breaker of
the power distribution circuit. If the ground of the device is open, a current
will flow through the patient to ground. The total resistance of the current
+
path is 25 + 25 0.5 kR = 50.5 kR.
The current passing through the patient is therefore = 120 = 2.4 mA.
50.5 kR
According to Table 12-1, the patient should feel the presence of the current.
Although this amount of current is not large enough to blow the fuse or trip
the circuit breaker, the level is not high enough to endanger the patient

Experimental work on dogs had shown that ventricular fibrillation could

be onset by a current as small as 20 microamperes (50-60 Hz) applied direct-
ly to the heart. Note that this current is 5,000 times below the "possible ven-
tricular fibrillation" current (100 mA according to Table 12-1) applied exter-
nally. In addition, it was shown that the threshold current triggering these
physiological effects increases with increasing frequency. For example, the
"let go" current increases from 10 mA to 90 mA when the frequency is
increased from 60 Hz to 10 kHz. Very high frequency current (e.g., 500 kHz)
used in electrosurgical procedures will not cause muscle or nerve stimula-
tion, and therefore will not cause ventricular fibrillation. Moreover, skin
bum and tissue damage can still occur at high current.

MACROSHOCK AND MICROSHOCK

The physiological effects described in Table 12-1 are often referred to as

macroshocks, whereas shocks that arise from current directly flowing
I

1
k
 Biomedical Device Technology: Princ@les and Design

through the heart are referred to as micro shocks. Figure 12-la and Figure
12-lb show the differences between macro and microshocks. In a
macroshock, the electrical contacts are at the skin surface. The risk current is
distributed through a large area of the patient's body. As shown in Figure
12-la, only a portion of the risk current flows through the heart. In a
microshock, the entire risk current is directed through the heart by an
indwelling catheter or conductor.

Figure 12-la. Macroshock.

Figure 12-lb. Microshock.

 Electrical Safty and Signal Isolation

Example 2
If the patient in Example 1 has a heart catheter (a conductor connected
directly to the heart) and it is connected to ground, calculate the risk current.

Solution
Since the catheter bypassed one skin-to-ground contact, the resistance of the
current path is now reduced to 25 + 0.5 k 0 = 25.5 k0. The risk current
therefore is equal to 120 = 4.7 rnA.
25.5 kR
Although this current level is still considered safe for external contact (Table
12-l), this amount of current under this situation (directly flowing through
the heart) will trigger ventricular fibrillation (>20 PA).

Table 12-2 shows the physiological effect of such current through the
heart. Note that this level of current is below the threshold of perception (list-
ed in Table 12-I), which suggests that the person will not be able to feel the
risk current even when it is sufficient to cause a microshock. Both macro-
shock and microshock can injure patients or cause death. Table 12-3 sum-
marizes the characteristics of macroshock and microshock.

Table 12-2.
-
Potential Hazards from Electrical Current (Cardiac Contact).
- - - - -- --

- -
Current Leuel -
Physiological Effect
- - -
-

0-10 p4 Safe for a normal heart

10-20 p4 Ventricular fibrillation may occur
20-800 pA -
Ventricular fibrillation
-- - -

Table 12-3.
Characteristics of Macroshock and Microshock.
-- - - -

Macroshock -
Microshock
-

Requires two contact points with Requires two contact points with
the electrical circuit at different electrical circuit at different
potentials potentials
High current passing through the Low current passing directly
body through the heart
Skin resistance is usually not Skin resistance is bypassed
bypassed, i.e., external skin contact
Usually due to equipment fault Usually due to leakage current
such as breakdown of insulation, from stray capacitors
exposure of live conductors, short
circuit
- .. -
-
of hot line to case
-- -
 Biomedical Device Technology: Princ$les and Design

The term leakage current is mentioned frequently in articles on patient

safety. Leakage current is not a result of a fault. It flows between any ener-
gized parts of an electrical circuit and its grounded parts. Leakage current
generally has two components-one capacitive and the other resistive.
Capacitive leakage current exists because any two conductors separated in
space have a certain amount of capacitance between the conductors. This
undesired capacitance is called stray capacitance. When an alternating volt-
age is applied between two conductors, a measurable amount of current will
flow (e.g., between the primary winding and the metal case of a power trans-
former, or between power conductors and the grounded chassis). The resis-
tive component of leakage current is primarily due to imperfect insulation
between conductors. Since no substance is a perfect insulator, some small
amount of current will flow between a live conductor and ground. However,
because of its relatively small magnitude compared to the capacitive leakage
current, resistive leakage current is often ignored.
In addition to electrical shocks and fire hazards described above, the loss
of electricity in health care settings can also create problems and compro-
mise the safety of patients. Table 12-4 summarizes different electrical haz-
ards in the health care environment.

Example 3
The total stray capacitance between the live conductors and ground of a
medical device powered by a 120 V, 60 Hz power supply is 0.22 pF. Find
the total capacitive leakage current flowing to ground.

Solution
Impedance due to the stray capacitance is

Therefore, magnitude of the capacitive leakage current is 120 = 10 PA.

12 M n

PREVENTION OF ELECTRICAL SAFETY HAZARD

Risk current is defined as any undesired current, including leakage cur-

rent that passes through the body of a patient. Although it cannot be avoid-
ed, it can be minimized by appropriate equipment deployment and proper
 Electrical Safety and Sipal Isolation

Table 12-4.
Summary of Electrical Hazards.
-- -- - -

Nature of Hazard
-

Macroshock Bums-including external (skin),internal, and cellular

Pain and muscle contraction-may cause physical injury
Ventricular fibrillation
Microshock Ventricular fibrillation
Fire and Explosion Damage and bum caused by heat or sparks from electrical short
circuit, overload in the presence of fuel, and enriched oxygen
environment
Electrical Failure Lost of function of life-supporting equipment
Disruption of service and treatment
Panic
- - --- - - --

utilization. Some simple user precautions that medical personnel can take to
ensure patient and staff safety are:
1. Medical personnel should ensure that all equipment is appropriate
for the desired application. Medical equipment usually has an
approval label on it that informs the operator of the risk level of the
equipment. One example of such classification and its meaning is
shown in Table 12-5. A patient leakage current is a current flowing
from the patient applied part through the patient to ground or from
the patient through the applied part to ground.
2. Ensure that the medical equipment is properly connected to an elec-
trical outlet that is part of a grounded electrical system. The power
ground will provide a low-resistance path for the leakage current.
3. Users of medical devices should be cautioned about any damage to
the equipment, including signs of physical damage, frayed power

Table 12-5.
Classification of Medical Electrical Equipment
(CAN/CSA C22.2 NO. 601-1 M90)
- -
- -

Maximum Allowable Patient

Leakage Current (jd)under
Equipment Intended Application Normal Single Fault
w e - - Conditions- - Conditions
B Equipment with causal patient contact, 100 500
usually have no patient applied parts
Equipment with patient applied parts
CF Equipment with cardiac applied parts, i.e.,
connected to the heart or to great vessels
leading to the heart
-- - - -
 Biomedical Device Technology: Princ$les and Design

cords, et cetera.
4. Ensure that there is an equipment management program in place so
that periodic inspections and quality assurance measures are per-
formed by qualified individuals to ensure equipment performance
and safety.
In addition to these user precautions, electrical systems and medical
devices can incorporate designs to lower the risk of electrical hazards. The
remainder of this chapter describes such designs.

GROUNDED AND ISOLATED POWER SYSTEMS

Figure 12-2 shows the line diagram of a common grounded electrical

system. In a grounded system, there are three conductors: the hot, the neu-
tral, and the ground (these wires are colored black, white, and green, respec-
tively, in a single phase 120 V North American power distribution system).
The neutral wire is connected to the ground wire at the incoming substation
or at the main distribution panel. In a 3-wire grounded system, the voltages
between the hot and neutral wires and between the hot and ground wires are
both 120 V. Under normal conditions, there is no voltage between the neu-
tral and ground conductors.
An isolated power system is shown in Figure 12-3. In this distribution
system, a line isolation transformer is placed between the power supply
transformer and the power outlets. In an isolated power system, there is no
neutral wire as both lines connected to the secondary of the transformer are
not tied to ground; they are floating with respect to ground (has no conduc-
tion path to ground). The voltage between the lines is 120 V. The color cod-
ing for the line 1, line 2, and ground conductor of an isolated power system
is brown, orange, and green, respectively. Both line conductors are protect-
ed by circuit breakers that are mechanically linked together.
In a grounded power distribution system, a short circuit between the hot
conductor to ground creates a large current to flow from the hot conductor
to ground. The spark and heat produced can create a fire. In an enriched
oxygen environment, an explosion may occur if flammable gas is present.
However, in an isolated power system, a short circuit to ground will not pro-
duce any significant current as none of the line conductors are connected to
ground (no return path exists for the fault current). As a result, fire and explo-
sion hazards due to ground faults are eliminated. Line isolation transformers
were required by building codes in operating rooms to prevent explosion
hazard when flammable anesthetic agents were used. Some jurisdictions
have dropped these requirements as flammable anesthetic agents are no
longer used in health care facilities.
 Electrical Safty and Signal Isolation

-Circuit breaker Hot

{ SUPPlY
transformer
To power
outlets
__+

Neutral

Ground

Figure 12-2. Grounded 3-Wire Power System.

Circuit breakers Line 1

n I
I
I
I
I
I
Isolation i
I
Supply I To power
transformer I
I outlets
I

i
Line 2
a

Ground

Figure 12-3. Isolated Power System.

Consider a ground fault (a hot conductor is connected to the grounded

chassis) occurring on a medical device plugged into a grounded power sys-
tem (Figure 12-4). If the ground conductor is intact, very little current will
flow through the patient; all current will be diverted through the ground con-
ductor to ground even when the patient is touching the chassis. The circuit
breaker in the hot wire will detect this excessive current and disconnect (or
 Biomedical Device Technology: Principles and Design

Ground fault

>->
Circuit breaker Hot
n
a
Circuit load of
medical device
supply
transformer

Neutral
m

Ground
>>
A
v >> -
- -
-
Figure 12-4. Ground Fault on 3-Wire Power System.

trip) the circuit. An intact ground connection is an effective first line of

defense against electrical shock.

Example 4
Consider the situation in which the chassis of the medical device in Figure
12-4 is not solidly grounded, if the ground fault current creates a 20 V poten-
tial difference between the chassis of the medical device to ground, calculate
the risk current when
i) The patient is touching the chassis and also touching a grounded object.
ii) The patient with a grounded heart catheter is touching the chassis.

Solution
i) Assuming the resistance of the current path is 50 kR, the risk current is
20 V/50 kR = 0.4 mA. This current is harmless and is not even notice-
able by the patient.
ii) Assuming the resistance of the current path is 25 kCl when one skin con-
tact resistance is bypassed by the catheter, the risk current is 20 V/25
kR = 0.8 mA. The microshock current will trigger ventricular fibrilla-
tion in the patient.
Normally, the circuit breaker will disconnect the device from power in a frac-
 Electrical Safety and Sipal Isolation

tion of a second to minimize patient injury. A grounded system with a pro-

tective circuit breaker or fuse is relatively safe to prevent electrical shock to
the patient when a ground fault happens. However, a spark may jump
between the hot and ground conductor at the fault location. Sparks may cre-
ate a fire or an explosion under the right situation.

Example 5
The ground connection of a medical device has a resistance of 1.0 0. If the
leakage current is 100 pA and a patient touching the grounded chassis has a
resistance to ground = 25 kR, find the risk current flowing through the
patient.

Solution

I ~~~~
The patient resistance is parallel to the ground
- connection resistance, the
current flowing through the patient resistance is I 4.0 = n ~ .
When the ground is intact, only 4 nA of current flows through the patient;
with a broken ground, the full leakage current (100 p4)will flow through the
patient.

For an isolated power system, a single ground fault between the line con-
ductors and ground will not produce a noticeable fault current as there is no
conduction path to complete the electrical circuit. However, if the fault is a
short circuit between one of the line conductors to the chassis of the medical
device, the chassis will become hot and therefore create a potential shock
hazard to the patient. A line isolation monitor (LIM) is used in an isolated
power system to detect this fault condition. A LIM is a device that monitors
the impedance of the line conductors to ground by periodically (several
times per second) connecting each of the line conductors to ground and mea-
suring the ground current (Figure 12-5). A LIM is usually set to sound an
alarm when the ground current exceeds 5 mA. Some alarms can be set from
1 to 10 mA. A LIM can detect a ground fault in an isolated power system as
well as deterioration of the insulation between line conductors and ground.
However, it cannot detect a broken ground conductor nor can it eliminate
microshock hazards.

Example 6
An isolated power system with a LIM set to sound an alarm at 5 mA is sup-
plying power to a patient location. The patient has a grounded heart catheter
 Biomedical Device Technology: Princ$les and Design

Circuit breakers Hot

n

Figure 12-5. Isolated Power System with Line Isolation Transformer.

and is touching another medical device. If the leakage impedance due to

capacitive coupling of the windings of the line isolation transformer is 25 kfl
and there is an insulation failure between a line conductor and the chassis of
the medical device (line shorted to metal chassis), what is the risk current to
the patient assuming the patient impedance is 30 kfl resistive?

Solution
The current flowing through the patient is equal to the line voltage divided
by the total impedance of the current path.
The magnitude of the risk current =
1 120 v
(30 kfl - j25 k n )
I= 3.0 mA.

The low level leakage current will not trigger the alarm of the LIM.
However, this will be a fatal current if it is allowed to flow directly through
the heart of the patient. In either case, the risk current would have been pre-
vented to flow through the patient if the equipment enclosure is properly
grounded.
 Electrical Safety and Signal Isolation

SIGNAL ISOLATION

Example 6 shows that an isolated power system with a line isolation

transformer is not sufficient to prevent microshock. In order to reduce
microshock hazard, medical equipment with patient applied parts in contact
with the heart or major blood vessels (F-type applied parts) are designed such
that the applied parts are electrically isolated from the power ground.
Isolation is achieved by using an isolation barrier with electrical impedance
of over 10 MR (such that the leakage current is lower than 10 FA). Figure
12-6 shows the block diagram of such a signal isolated device. The compo-
nents enclosed by the dotted line are electrically isolated from the power
ground. Note that in order to achieve total isolation, the power supplies to the
components before the signal isolation barrier will also need to be isolated.
There are two different ground references in an isolated patient applied part
medical device. The ground reference for the nonisolated part of the circuit
are connected to the power ground, while the ground reference for the iso-
lated components are connected to the ground reference of the isolated
power supply. These two grounding references are not connected together
and are differentiated by two different grounding symbols (Figure 12-6).

I Isolated circuit I
I
I
I
I
i\mpl~fier 1 ~\rnpl~lter
Transducer + and signal
condtttoner
+ Siwal
rso~tlon
I
-b and *lgn"
processor
+ Output

t t I

II
I
-
-
A
-
-
A

L----- -, ---------I

-L 4
Powcr
gm~md
Is~>laled
ground
t -
Power
SUPPlY

-
-
Figure 12-6. Block Diagram of Mehcal Device with Signal Isolation.

Signal isolation is achieved by an isolation barrier. Common isolation

barriers used in medical instrumentation are isolation transformers or optical
isolators (Figure 12-7). Isolation transformers used in signal isolation are
much smaller in size than those used in power system isolation as they do not
 Biomedical Device Technology: Princ$les and Design

need to transform high power. Furthermore, they have much higher isolation
impedances. Optical isolators break the electrical conduction path by using
light to transmit the signal through an optical path. Figure 12-7a shows a sim-
ple optical isolator using a light-emitting diode (LED) and a phototransistor.
The signal applied to the LED turns the LED on at high voltage level and off
at low voltage level. The phototransistor is turned on and off according to the
light coming from the LED. Since low-frequency signals often suffer from
distortion when passing through isolation barriers, a physiological signal is
first modulated with a high-frequency carrier ( e g , 50 kHz) before being sent
through the isolation barrier. A demodulator removes the carrier and
restores the signal to its original form on the other side of the isolation bar-
rier. The signal at the output of the isolation circuit should be the same as the
signal at the input.

r------I
I I

Input Modulator Demodulator

Outp$
signal signal
I I
L------J
(a) Optical Isolator

'
3 1-1
I
I
I
lnput Modulator I I Demodul~
I
signal I
I I
I-------]

(b) Transformer Isolator

Figure 12-7. Optical and Transformer Signal Isolation.

OTHER METHODS TO REDUCE ELECTRICAL HAZARD

Equipotential Grounding
For a patient with the skin impedance bypassed, a tiny voltage can cre-
ate a microshock hazard. For example, with a current path impedance of 2
 Electrical Safety and Signal Isolation

kln, a voltage difference of 20 mV is sufficient to cause a 10 pA patient risk

current. For this reason, it is desirable to protect electrically susceptible
patients by keeping all exposed conductive surfaces and receptacle grounds
in the patient's environment at the same potential. It is achieved by con-
necting all ground conductors (equipment cases, bed rail, water pipes, med-
ical gas outlets, etc.) in the patient's immediate environment together and
making common ground distribution points in close proximity to patients.

Ground Fault Circuit Interrupters

For an electrical device, the current flowing into the hot conductor
should be equal to the current in the neutral conductor. When there is a cur-
rent flowing from the hot conductor to ground, such as the existence of leak-
age current or a ground fault, the current in the hot conductor is different
from that in the neutral conductor. A ground fault circuit interrupter (GFCI)
senses the difference between these two currents and interrupts the power
when this difference, which must be flowing to ground, exceeds a fixed value
(e.g., 6 mA). Figure 12-8 shows the principle of a GFCI. Under normal con-
ditions, there is no magnetic flux in the sensing coil as the hot and neutral
current are equal. When there is a large enough hot to ground current, the

Line

Figure 12-8. Ground Fault Circuit Interrupter.

net magnetic flux will trip the circuit breaker.

If a person is touching a hot conductor with one hand and a grounded
object with the other, a risk current will flow from the hot conductor through
 Biomedical Device Technology: Plinc$les and Design

the patient to ground. The GFCI detects the current difference in the hot and
neutral conductors and interrupts the power before it becomes a problem.
This protects the person from macroshock. GFCIs are commonly used in
wet locations where water increases the electrical shock hazard. However, a
GFCI is not used in critical patient care areas (OR, ICU, CCU) where life
support equipment may be in use as it may be too sensitive to cause unnec-
essary power interruption.

Double Insulation
A device with double insulation has an additional protective layer of
insulation to ensure that the outside casing has a very high value of resistance
or impedance from ground. This separate layer of insulation prevents con-
tact of any person with any exposed conductive surface. Usually the outside
casing of the equipment is made of nonconductive material such as plastic.
Any exposed metal parts are separated from the conductive main body by
the addition of a protective, reinforcing layer of insulation. All switch levers
and control shafts must also be double insulated (e.g., using plastic knobs). In
order to be acceptable for medical equipment, the outer casing must be
waterproof, that is, both layers of insulation should remain effective, even
when there are spillages of conductive fluids. Double insulated equipment
need not be grounded, so its supply cord does not have a ground pin.

Batteries and Extra-Low-VoltagePower Supply

The higher the power supply voltage, the higher the current that will flow
through a person in an electrical accident. To reduce the voltage, one can use
a step-down transformer or a low-voltage battery. In general, a voltage not
exceeding 30 V r m s is considered as extra-low voltage. Such voltage is safe to
touch for a healthy person. However, excessive heat from a direct short cir-
cuit (e.g., a battery short circuit), can create burn or explosion hazards.
Battery-powered equipment is very common in health care settings because
in addition to its lower electrical risk, it provides mobility to the equipment
as well as the patient. A common type of equipment using a battery as the
main source of power is the infusion pump.
Table 12-6 summarizes the effectiveness of different electrical safety pro-
tection measures discussed above toward microshock and macroshock.
 Electrical Safety and Signal Isolation

Table 12-6.
Summary of Shock Prevention Methods
-- - - - -

Macroshock Microshock
Proper Grounding Yes Yes
Double Insulation Yes Yes
Isolated Power System Yes No
Isolated Power with LIM Yes No
Isolated Patient Applied Parls Yes Yes
Equipotential Grounding N/A Yes
Ground Fault Circuit Interrupter Yes No
Battery Powered Yes Yes
Extra-Low Voltage (AC) Yes No

MEASUREMENT OF LEAKAGE CURRENT

It was mentioned earlier that the levels of electrical shock threshold cur-
rent increase with increasing frequency. Figure 12-9 shows the approximate
relationships between the threshold current and the power frequency. The
higher the frequency, the higher the risk current to trigger physiological
effects. To take into account the frequency-dependent characteristics of the
human body's response to risk current, a measurement device to measure
device leakage current is specified in the International Electrotechnical
Commission standard (IEC601-1) on electrical safety testing of medical
devices and systems. The measurement device consists of a passive network

Normalized

t threshold current

Frequency (kHz)

0.1 1 10 100 1,000

Figure 12-9. Threshold Electric Shock Current Versus Frequency.
 Biomedical Device Technology: Princ$les and Design

with a true RMS millivoltmeter, which essentially simulates the impedance

of the human body as well as the frequency-dependent characteristics of the
body to risk current. Figure 12-10 shows the IEC601-1 leakage current mea-

IL

Voltmeter
1 klZ 0.015 pF IL

-
m
-
m

Figure 12-10. IECGO1-1 Leakage Current Measurement Device.

surement device.
If the leakage current flowing into the measurement device is IL,the cur-
rent flowing through the capacitor I1 is equal to

where Xc = 1 a.
j2nf X 0.015 X lo-"
The voltage across the capacitance as measured by the voltmeter is equal to

For a direct current, f = 0 and Xc = a,therefore, Vm = lW IL.

From this relationship, if the voltmeter is measuring 1 mV, the leakage
current is equal to 1 pA. Therefore, if the voltmeter is set to read in mV, the
value displayed is the leakage current in pA.
For low-frequency current such as 60 Hz power frequency, Xc = -j 177 X
lo3a , therefore
 Electrical Safety and Signal Isolation

However, when f is very large, Xc 4 0, therefore,

This analysis illustrates that the IECGOl-1 leakage current measurement

device has an inverse characteristic to the threshold electrical shock response
(Figure 12-9). Instead of having different threshold current levels for differ-
ent leakage current frequencies (e.g., 10 pA at 60 Hz and 90 pA at 10 kHz),
one can fix the threshold current limits and use the IECGOl-1 measurement
device to compensate for the frequency response (e.g., 10 pA maximum
allowable leakage current for all frequencies). Figure 12-11 shows how the

Medical
device
under test

Figure 12-11. Use of IEC601-1 Measurement Device for Leakage Current.

measurement device is used to measure the patient leakage current flowing

from the device through a patient applied parts to ground.
Interested readers should refer to the Standards (e.g., IEC60601) for the
leakage current limits and the details of how and what to measure.
 Chapter 13

MEDICAL WAVEFORM DISPLAY SYSTEMS

OBJECTIVES

State the functions and characteristics of medical chart recorders and

displays.
Identify the basic building blocks of a paper chart recorder.
Explain the construction of mechanical stylus and thermal dot array
recorders.
Explain the principles of operation of laser printers and ink-jet printers.
Explain the terms non-fade, waveformparade, and erase bar in medical dis-
plays.
Compare and explain the principles of different types of medical dis-
plays.
Analyze the performance characteristics of medical display systems.

CHAPTER CONTENTS

1. Introduction
2. Paper Chart Recorders
3. Visual Display Monitors
4. Performance Characteristics of Display Systems

INTRODUCTION

For a medical device, the output device is the interface between the
device and its users. Some common output devices found in medical equip-
 Medical Waueform Display Systems

ment are listed in Table 1-3 of Chapter 1. They include paper records, audi-
ble alarms, visual displays, et cetera. A video monitor that displays a medical
waveform such as an electrocardiogram is a typical medical output device.
The principles of paper chart recorders and video display monitors are dis-
cussed in this chapter.

PAPER CHART RECORDERS

The function of a paper chart recorder in a medical device system is to

produce records of physiological waveforms and parameters. These records
can be used:
as a snapshot of the patient's physiological condition for future refer-
ence, or
to record an alarm condition so that it can be reviewed later by a med-
ical professional.
Charts are often considered as medical records and therefore are
required to be stored for a long period of time (e.g., 7 years for an adult
patient in Canada). A paper chart recorder may use ink (e.g., ink stylus, ink-
jet) or heat (e.g., thermal stylus, thermal dot array) to produce the waveform
on a piece of paper. In an ink recorder, ink from an ink reservoir is supplied
to a writing mechanism to produce the waveform on a piece of paper.
Recorders using thermal styli or thermal dot array print heads require the
use of heat-sensitive or thermal papers. Unlike ordinary paper, thermal
paper is coated with a special chemical that turns dark when heat is applied.
Thermal paper is more expensive than ordinary paper. However, the design
and maintenance of thermal writers are simpIer than those of ink writers.
Paper chart recorders can be divided into two categories: continuous
paper feed recorders and single page recorders. Thermal dot array print
heads are commonly used in continuous paper feed recorders. Lasers print-
ers are in general the recorder of choice for single page recorders. Examples
of paper chart recorders are listed in Table 13-1.

Table 13-1.
-
Paper Chart
- -
Recorders.
Continuous Paper Feed Mechanical stylus recorder; thermal dot array recorder
Single Page Feed Ink-jet printer; laser printer
- - - -- --
 Biomedical Device Technology: Princ$les and Design

Continuous Paper Feed Recorders

A continuous feed paper recorder draws paper from a paper roll or a
stack of fanfolded paper. It can continuously record a waveform for an
extended period of time. Two types of continuous paper feed chart recorders
are found in medical devices: the mechanical stylus recorders and the ther-
mal dot array recorders.

Mechanical Stylus Recorders

A mechanical stylus recorder consists of an electromechanical transduc-
er to convert the analog electrical signal (e.g., amplified biopotential signal)
to a mechanical motion. This motion is mechanically linked to a writing
device such as an ink pen or a thermal stylus. The writing device leaves a
trace on the chart paper as it moves across the paper. The paper is fed from
a paper chart assembly that includes a paper supply mechanism and a writ-
ing table. The paper supply is driven by a motorized paper driving mecha-
nism. Figure 13-1 shows a paper chart recorder setup using a galvanometer
as the electromechanical transducer. A servomotor drive can replace the gal-
vanometer to drive the mechanical stylus.

Wrltlng stylus
Mechanical

Galvanometer

Paper chart D~rect~on

of paper chart movement
Figure 13-1. Mechanical Stylus Paper chart Recorder.

Thermal Dot Array Recorders

In a thermal dot array recorder, the electromechanical transducer is
replaced by a thermal dot array printhead. The printhead of a thermal dot
 Medical Waveform Display Systems

array recorder consists of a row of heater elements placed on top of the mov-
ing thermal paper as shown in Figure 13-2. Each of these heater elements
can be independently activated to leave a black dot on the heat-sensitive
paper. The analog electrical signal is first converted to a digital signal and
then processed to heat the appropriate dots in the printhead. The paper sup-
ply and drive assembly is similar to that of the mechanical stylus recorder.
Movement of the paper in conjunction with the appropriate addressing of
the thermal elements on the printhead produces the image on the paper. As
there are only a finite number of thermal elements on the printhead, the
trace recorded on the paper is not continuous like the trace produced by a
mechanical stylus writer. The vertical resolution of the recorder is limited by
the number of thermal elements in the printhead. Paper chart recorders used
in physiological monitors usually have resolution better than 200 dots per
inch (dpi). Table 13-2 lists the major functional components of the mechan-
ical stylus recorder and the thermal dot array recorder.
Mechanical stylus chart recorders are being replaced by thermal dot
array recorders in medical devices as they have fewer mechanical moving
parts. Mechanical stylus recorders require a higher level of maintenance due
to wear and tear and misalignment problems.

Dot matrix Thermal dots

Paper

Paper supply

Paper chart
/ Direction of paper chart movement
Figure 13-2. Thermal Dot Array Paper Chart Recorder.

Single Page Feed Recorders

Single page feed recorders use standard size paper (such as 8.5" by 11"
paper) and therefore can record only a finite duration of the waveform on a
single sheet of paper. Off-the-shelf laser printers or ink-jet printers are com-
monly used. Figure 13-3 shows the construction of a laser printer.
 Biomedical Device Technology: Principles and Design

Table 13-2.
Main Building Blocks of Continuous Paper Feed Chart Recorders.
- - - -- -

Mechanical Stylus Recorder

- --
Thermal Dot Array (Dot Matrix) Recorders
- - - - - - -

An electromechanical device The analog electrical signal is converted

(galvanometric or servomotor drive) to to digital signal via an A/D converter.
convert electrical input signal to
mechanical movement.
An arm to transmit the mechanical The digital signal is being processed to heat
movement to the stylus. the appropriate dots in the dot matrix

A stylus to leave a record of the signal on The activated dot in the print head will
the chart paper as it moves across the paper. leave a black dot on the heat-sensitive paper.
It can be an ink stylus or a thermal stylus.
A chart paper assembly consisting of a paper supply mechanism and a paper writing table.
A paper drive mechanism to move the chart paper across the table.
- -

Time varying
Scanning
mirror

Figure 133a. Laser Printer Writer System.

Laser Printers
For a laser printer, the time varying signal such as an ECG is first con-
verted to digital signal by an analog to digital converter. A photosensitive
drum in the printer rotates at a constant speed. The speed of rotation of the
drum determines the paper speed of the chart. The rotational motion of the
scanning mirror reflects the laser beam to move across the surface of the
 Medical Waveform Display Systems

Prlmaty corona wire (negative charge)

Resin particles (toner)
\ Laser beam

Paper
move to
fuser

Transfer corona wire (poslive charge)

Figure 13-3b. Laser Printer Printing System.

drum (Figure 13-3a). The laser diode is switched on and off by the print
processor according to the ECG signal and the position of the scanning mir-
ror. The section of the rotating photosensitive drum acquires a negative
charge when it passes by the primary corona wire (Figure 13-3b). When the
laser beam reflected by the scanning mirror strikes the spots on the drum
where dots are to be printed, the spots on the surface of the negatively
charged drum become electrically neutral. As the drum rotates, new rows of
neutral spots form in response to the laser pulses.
The surface of the developing cylinder contains a weak magnet field. As
the developing cylinder rotates, it attracts a coating of dark resin particles
(toner) that contain bits of negatively charged ferrite. As the resin particles
on the developing cylinder move closer to the photosensitive drum, these
particles, due to their negative charge, are repelled by the negative charged
area on the drum and moved to the neutral spots that were created earlier by
the laser beam.
As the resin particles on the drum move toward the paper, they are being
attracted to the paper by the positive charge on the paper created by the
transfer corona wire. The image on the drum is therefore transferred onto
the paper.
As the drum continues to rotate, a cleaning blade removes all residue
resin particles on the drum surface and an erase lamp introduces a fresh neg-
ative charge uniformly on the entire surface of the drum. This prepares the
drum to receive the next part of the page information.
To fuc the image on the paper, the resin particles are heat-fused onto the
 Biomedical Device Technology: Principles and Design

paper and the charge on the paper is neutralized by passing the paper over
a grounded wire brush. Laser printers have resolution of 600 dpi or higher.

Ink-Jet Printers
An ink-jet printer has a printhead with multiple print elements as shown
in Figure 13-4a. Each element consists of a tiny aperture with a heat trans-
ducer behind it. When activated, the heater boils the ink and forms a vapor
bubble behind the aperture. The vapor pressure forces a minute drop of ink
out toward the printing surface to create a single image dot.
The printhead is coupled to an ink reservoir to form the print cartridge.
The print cartridge is driven by a servomotor to move back and forth across
the paper. Together with the translational motion of the paper, a time vary-
ing waveform can be recorded on the paper (Figure 13-4b). Instead of using
heat to create the ink-jet, some printers use the mechanical vibration force
created by a piezoelectric crystal to eject the ink onto the paper.

Ink compartment
in printhead

Ink droplet

Heat transducer

VISUAL DISPLAY MONITORS

In a physiological monitoring system, a visual display monitor provides

real-time visual information for the medical professional to assess the condi-
tion of the patient. Some common monitors used in medical applications are:
 Medical Waveform Display Systems

............ .

of paper chart movement

D~rect~on

Figure 13-4b. Ink-Jet Paper Chart Recorder.

Cathode ray tube (CRT)

Liquid crystal display (LCD)
Plasma display
Electroluminescent display

Cathode Ray 'hbe Displays

The cathode ray tube has been used for a long time as an efficient and
reliable visual display monitor for medical devices. Figure 13-5a shows the
main components of a CRT. It is also called an oscilloscope.
The electron gun provides a continuous supply of electrons by heating
the cathode filament (thermal ionic effect).These electrons are attracted and
accelerated toward the screen due to the high voltage (several kV) applied
across the anode and cathode. When the high-velocity electron beam hits the
phosphor screen, it emits visible light photons at the location of impact. To
deflect the beam to a desired location on the screen, a voltage is applied
across two pairs of deflection plates located above, below, and on each side
of the electron beam. In normal time-base operations, a saw-tooth waveform
voltage is applied across the horizontal deflection plates. If no signal is
applied across the vertical deflection plates, this saw-tooth voltage moves the
electron beam horizontally back and forth across the tube. When the time
varying waveform to be displayed is amplified and applied across the verti-
cal deflection plates, it causes the electron beam to move in the vertical direc-
tion according to the voltage level of the signal. When the signal frequency
 I
-
H~ghvoltage DC
-
Biomedical Device Technology: Princ$les and Design

Electron beam

\
+

I
\

-........{..+.+...........
I l l
I l l
I
....................
..................
lil 4 - / /
filament
I
Electron Control and Hor~zontaldeflect~on Vertical deflect~on
focus~ng plates plates
OUn gr~ds

Accelerating anode

r'igure 13-5a. Basic Structure of Cathode Ray Tube.

is a multiple of the frequency of the saw-tooth waveform, a steady waveform

of the signal is displayed on the screen of the CRT (Figure 13-5b).
Adjusting the frequency of the saw-tooth waveform so that a steady sig-
nal is displayed on the CRT screen is called triggering. Automatic triggering
is done by feeding the signal to be displayed to a frequency detector and
using this frequency to generate the frequency of the saw-tooth horizontal

Vertical deflection plate voltage

I Time

t Horizontal deflection plate voltage

fl Time

Figure 13-5b. Cathode Ray Tube Display

CRT screen
 Medical Waveform Display Systems

deflection signal. The amplitude of the signal displayed on the screen can be
adjusted by changing the amplification factor (sensitivity control) of the
amplifier feeding the vertical deflection plates.
In order to control the brightness of the display and the convergence of
the electron beam to a small dot when it reaches the phosphor screen, con-
trol voltages are applied to a set of control and focusing grids in front of the
electron gun. Instead of using electrostatic deflection plates, some oscillo-
scopes use electromagnetic coils to provide horizontal and vertical deflec-
tions to the electron beam.
For a fast-moving repetitive waveform, the persistence of the phosphor
and the response time of the human eye will make a triggered waveform
appear to be continuous and stationary on the screen. For nonperiodic wave-
forms, as each sweep (one cycle of the saw-tooth waveform) produces a dif-
ferent trace of waveform on the screen, no stationary waveform will be seen.
For slow time varying signals, since the phosphor can scintillate only for a
fraction of a second, the trace will appear as a dot moving up and down
across the screen (this is why old physiological monitors were referred to as
bouncing ball oscilloscopes). To produce a steady display even from nonpe-
riodic and slow varying signals, a storage oscilloscope is required. Figure
13-6 shows the block diagram of a storage oscilloscope. Instead of sending
the signal to be displayed directly to the vertical deflection plates, the signal
is first converted to digital format by an analog to digital converter (ADC)
and stored in the memory. This slow varying signal is then reconstructed and
swept across the screen many times faster than its original frequency so that
it can be seen as a solid trace on the screen. This category of display is called
non-fade displays.
There are two types of non-fade displays: one is "waveform parade" and
the other is "erase bar." In a waveform parade non-fade display, the wave-

To vertical
deflection
AID DIA
input
Amplifier + converter -b b converter b

t t
Binary D/A
counter -b converter . b
Saw-tooth
A waveform
to horizontal
deflection
-b Display
control -

Figure 13-6. Storage Oscilloscope.

 Biomedical Device Technology: Principles and Design

Newest data Oldest data

\ I

(a) Waveform Parade (b) Erase Bar

Figure 13-7. Non-fade Display.

form appears to be moving across the screen with the newest data coming
out from the right-hand side of the screen and the oldest data disappearing
into the left-hand side (Figure 13-7a). In an "erase bar" non-fade display, the
data appear to be stationary. A cursor (or a line) sweeps across the screen
from left to right (Figure 13-7b). The newest data emerges from the left-hand
side of the cursor while the oldest data are erased as the cursor moves over
them. When the cursor has reached the right edge of the screen, it disappears
and then reappears from the left edge of the screen.
CRT displays are bright, have good contrast ratio (ratio of output light
intensity between total bright and dark), high resolution, and high refreshing
rate. However, they are heavy, bulky, and may have uneven resolution
across the screen.

Liquid Crystal Displays

Liquid crystal displays (LCDs) have gained popularity to replace CRTs
as the display of choice for medical devices in recent years. LCDs are light-
weight, compact, and robust compared to CRTs. LCDs operate under the
principle of light polarization. A typical liquid crystal cell is shown in Figure
13-8. Liquid crystal, which has the ability to rotate polarized light, is sand-
wiched between two transparent electrodes and two polarizers with axes
aligned in the same direction to each other. The light from a light source at
the back of the cell is polarized by the first polarizer before it passes through
the liquid crystal and then to the other polarizer. When no voltage is applied
across the electrodes, the axis of the polarized light is rotated (or twisted) 90'
 Medical Waueform Display Systems

by the liquid crystals. As the axis of the analyzing polarizer is in line with that
of the polarizing polarizer, the polarized light is blocked and therefore no
light will exit from the other end. When a voltage (about 5 to 20 V) is applied
across the electrodes, the twisting effect of the liquid crystal disappears.
Since the axis of the polarized light is in line with the axis of the analyzing
polarizer, the polarized light can exit through the other end with little atten-
uation. By switching the voltage across the electrodes, the liquid crystal cell
can be turned on (bright)or off (dark).This is called a twisted nematic (TN)
LCD.

Liquid cfysteJ with electrodes

off
dark

Polarizing polarizer Analyzing polarizer

Light
4 4 4
-p, bflght
O"
I
1
I I I I

I I

Figure 13-8. Principle of Operation of Liquid Crystal ;tli~Ly.

To illustrate the operation of a two-dimensional display, a 5 pixel x 5

pixel LCD panel with column electrodes and a polarizer on one side plus
row electrodes and a polarizer on the other side of the LCD crystal is shown
in Figure 13-9. Each of the 25 pixels sandwiched between the horizontal and
vertical addressing electrodes can be turned on and off by applying appro-
priate voltages to the rows and column electrodes. For example, if we want
to turn on the pixel B-3, the column electrode B will be connected to a pos-
itive voltage and the row electrode 3 will be connected to a negative voltage.
To display a time varying signal, a display driver converts the input signal to
be displayed into column and row addressing sequences to turn the pixels on
and off.
The brightness of a pixel is controlled by adjusting the "on" time or the
applied voltage across the liquid crystal. The longer the duty cycle of the
 Biomedical Device Technology: Principles and Design

Figure 13-9. LCD Panel Addressing.

applied voltage, the brighter the pixel appears to be. To add color to the dis-
play, primary color filters (red, green, and blue) are overlaid on top of the
pixel elements. Multiple colors are created by combining different intensities
of these primary colors. As it requires three pixels to form one color pixel, a
color LCD display will require three times as much pixel to achieve the same
resolution as a monochromatic monitor. A 640 X 480 color LCD panel
requires 920,000 LCD pixel elements.
For this type of LCD display, the addressing frequency and hence the
screen refreshing rate is limited by the capacitance formed by the addressing
electrodes and the LCD crystal (two conductors separated by an insulator).
Earlier LCD panels using passive addressing electrodes suffered from slow
refreshing rate and often showed a "tail" following a fast-moving object.
Employing thin film transistors (TIFT)to form an active matrix (AM) has sub-
stantially increased the screen refreshing rate in modern LCD displays.
However, since each pixel element requires one T l T , AMLCD panels are
more expensive than passive LCD panels. Figure 13-10 shows the schemat-
ic diagram of a single pixel element of an AMLCD. For a 640 X 480 color
display, 920,000 TIFT are required to be fabricated on a single substrate.
A LCD does not emit light photons. An external light source is therefore
required to display images on a LCD. A back-lit LCD has a light source
Iocated at the back of the LCD (Figure 13-8). A reflective LCD uses a mir-
ror at the back to reflect light coming from the front, such as daylight or
ambient light. Other than its slower screen refreshing rate, a LCD has lower
 Medical Wau$oorm Display Systems

Column addressing
electrode--SoURCE Thin film transistor

Row addressing
electrode--GATE

Liquid crystal cell

Figure 13-10. Schematic Diagram of an AMLCD Pixel.

contrast ratio and narrower viewing angle (brightness of the LCD is lower
when viewing from the sides and above or below the display) than a CRT.

Plasma Panel Displays

Plasma panels used in low-resolution and alphanumeric displays are
established technology and have been proven to be rugged and reliable. In
recent years, high-resolution large flat panel plasma displays are used in con-
sumer products such as televisions as well as medical applications. Plasma is
a gas made up of free-flowing ions and electrons. When an electrical current
is running through plasma, it creates a rapid flow of charged particles collid-
ing with each other. These collisions excite the gas atoms, causing them to
release energy in the form of photons. In xenon-neon plasma, most of these
photons are in the ultraviolet region, which is invisible to the human eye.
However, these ultraviolet photons can interact with a phosphor material to
produce visible light.
The cross-sectional view of a plasma display cell is shown in Figure
13-1 1. The cell is sandwiched between two glass plates. The addressing elec-
trodes are surrounded by a dielectric insulating material covered by a mag-
nesium oxide protective layer. The plasma is trapped in an enclosure coated
with phosphor.
In a plasma flat panel display, each pixel is made up of three separate
plasma cells or sub-pixels with phosphors chosen to produce red, green, and
blue light. Different colors can be produced by combining different intensi-
 Biomedical Device Technology: Principles and Design

ties of these primary colors by varying the current pulses flowing through
each of the three sub-pixels. Similar to a LCD, row and column addressing
electrodes are used to produce the image. Plasma panels are less bulky than
a CRT, and they have a higher contrast ratio and higher response rate than
a LCD. However, they require higher driving voltage than a LCD (150-200 V).

D~electrtcthin film Row electrodes

Xenon-neon plasma
\ /

/
Glass plate

Phosphor-coated
cell Interior

I / I
I
I
I
I
I
I
I
I
I
I
I
Power source

Magnes~umoxide
protechve layer
t Visible
i il~ghtioutput
i i Transparent .column
electrodes
Figure 13-1 1. Basic Construction of a Plasma Display Cell.

Electroluminescent Displays

Electroluminescent (EL) displays contain a substance (such as doped zinc

sulfide) that produces visible light when an electric field is applied across it.
Figure 13-12 shows an electroluminescent panel with the electroluminescent
material between rows and columns of addressing electrodes. Electrolumi-
nescent displays are thin and compact, with high response rate, and accept-
able brightness. However, they are not as popular than LCD in medical
applications as they require relatively high driving voltage (170-200 V) and
have lower power efficiency.
Table 13-3 shows a general comparison of the three common display
technologies.

PERFORMANCE CHARACTERISTICS OF DISPLAY SYSTEMS

The output of a display system is often used in diagnosis or to monitor

the condition of a patient during medical treatment. The accuracy of the dis-
 Medical Waveform Display S y s t m

Doped zinc sulfide light Row electrodes

Dielectric thin film emitting thin-film , I

Glass

Figure 13-12. Basic Construction of Electroluminescent Display.

Table 13-3.
- - -
Comparison Chart on Display Technologies
- - - - -- - - -

- - - -
CRT -
LCD-
Plasma
- - - -
EL -

BRIGHTNESS High Need High High

light source
High Low High High
Washout Yes Washout Washout

Good Good but Good but Good but

expensive expensive expensive
Wide Narrow Wide Wide
High Low High High
Heavy Light Medium Medium
Bulky Thin Thin Thin
DRIVING
VOLTAGE High Low Medium Medium
COST Low High High High
(for AMLCD)
- - - - -- - - - - - -

play can therefore affect the medical outcome. This section studies the per-
formance characteristics of the paper chart recorders and the video display
monitors discussed earlier in this chapter. Most of the medical device per-
formance characteristics and parameters discussed in Chapter 2 apply to the
display systems. Some of them are discussed below.
 Biomedical Device Technology: Principles and Design

Sensitivity
The function of a paper chart recorder as well as a video display moni-
tor is to convert the electrical input signal (usually a voltage signal) to a ver-
tical deflection (distance). The sensitivity therefore is usually measured in dis-
tance per unit voltage. For example, the vertical sensitivity of an electrocar-
diograph (ECG)may be 5 mm/mV, 10 mm/mV, or 20 mm/mV. Many med-
ical devices with an output display have an internal calibration signal to
enable users to quickly verify the accuracy of the sensitivity. An example is
the 1 mV internal calibration square pulse of an electrocardiograph. When
invoked, the size of the square pulse will be shown according to the sensitiv-
ity setting. For example, when set at 10 mm/mV, a square pulse with 10 mm
amplitude will be recorded or displayed. An external calibration signal can
also be applied to the input to verify the accuracy of the display.

Paper/Sweep Speed
A physiological signal is often a time varying signal. The distance on the
horizontal axis of the display represents the elapsed time of the signal. For an
ECG monitor, a common paper speed of the recorder and sweep speed of
the monitor is 25 mm/sec. To check the accuracy of the paper speed or
sweep speed, a repetitive signal of known frequency is applied to the input
of the display device. The horizontal distance of one cycle of the output sig-
nal is measured. The sweep speed of the display is equal to this distance
divided by the period of the applied signal.

Resolution
Resolution of a display is a measure of the smallest distinguishable
dimension of an image on the display. For a paper chart recorder using a
thermal dot array, the resolution may be 8 dots/mm in the vertical direction
and 32 dots in the horizontal direction. For a video monitor, the resolution
may be expressed as 1,280 X 800 pixels. If the dimensions of this monitor
are 40 cm X 22.5 cm, the display resolution is 32 pixels/cm horizontally and
20 pixels/cm vertically.

Frequency Response
Like any transducers and functional components, a display of a medical
device has limited bandwidth. A typical frequency response of a display sys-
tem is shown in Figure 13-13. An ideal display system should have a trans-
 Medical Waveform Display Systems

fer function with the lower cutoff frequency (fi) lower than that of the signal
that is going to be displayed and an upper cutoff frequency @) higher than
that of the signal. The regions of the transfer function between the upper and
lower cutoff frequency should be constant or flat.

Example 7
An electrocardiogram is shown. If the sensitivity and paper speed settings are
10 mm/mV and 25 mrn/sec, respectively, find the amplitude of the ECG sig-
nal and the patient's heart rate. (Note: One small square on the chart equals
1 mm).

Solution
The amplitude of the R wave is 10 mm. As the sensitivity setting is 10
mm/mV, the input ECG signal amplitude is 1 mV.
The distance between two QRS complexes is 25 mm (one cycle). With a
paper speed of 25 mm/sec, this represents a period of 1 second. Therefore,
the heart rate is 60 beats per minute.

The frequency response transfer function of the display system can be

obtained by inputting a sinusoidal signal of known amplitude and frequency
and measuring the amplitude of the output. By changing the input frequen-
cy and repeating the measurement, the frequency response transfer function
can be plotted. Using this method to obtain the lower cutoff frequencies of a
low-frequency signal such as an electrocardiograph (fi = 0.01 Hz) can be dif-
ficult and time-consuming. In practice, the lower cutoff frequency of the dis-
play (or any system) can quickly be estimated by assuming that the display
 Biomedical Device Technology: Principles and Design

-
Medical Vo(f)
Vl(f)
display

Figure 13-13. Frequency Response of a Medical Display System.

behaves like a first-order high pass filter. To find the cutoff frequency of the
high pass RC filter as shown in Figure 13-14, a step function is applied to the
input to produce the step response from the output. From the step response,
fican be obtained from the equation:

where the exponential decay time constant RC can be obtained by using the
equation:

- First-order high pass filter

I :
Input signal (step input)

I t2 t1 7
Output signal (step response)

Figure 13-14. Estimation of Display Lower Cutoff Frequency.

 Medical Waveform Display Systems

Example 2
Find the lower cutoff frequency of a paper chart recorder if the step response
is as shown in Figure 13-14.

Solution

--
RC
Using the exponential decay equation Vo = Vie , at time tr and t2,
-- t 1 -- k
RC RC
VI = Vie ,and V2 = Vie ,dividing the first equation by the second gives
--. ,
Vz=e RC
+RC=- tl - t2 . But f ~ = - , so the lower cutoff frequency
v v
In -2 2nRC
v1
of the chart recorder can be calculated by looking up the voltage Vr and Vi
at time tr and h from step response.
Alternatively, if we pick t2 as the start of the step input and tr is the time when
the step response dropped to 50% of the initial value, from the above-derived
equation

tr is the time elapsed when the output decreases to 50% of the initial value.

Example 3
The following chart paper recording was made by an electrocardiograph in
response to a step input. Estimate the lower cutoff frequency$ of the unit if
the paper speed is 25 mm/s. The vertical axis of the chart is 1 mV/div. and
the horizontal axis is 5 mm/div.
 Biomedical Device Technology: Principles and Design

Solution
The ,5O% amplitude of the output is seven divisions from the beginning of the
itep input. Assuming the response is a first-order high pass filter. Using the
equation
f~ = -
0'11 derived above, t~ = 7 x 5 m m = 1 . 4 ~
tl 25 mm/s
0 11 -
Therefore, f~ = - --0.11 - 0.079 H~
tl 1.4
 Chapter 14

PHYSIOLOGICAL MONITORING SYSTEMS

State the functions of a physiological monitor.

Sketch the functional block diagram of a typical multiparameter patient
monitor.
* List the common physiological parameters being monitored.
Identify the common features of a bedside, ambulatory, and central
monitor.
Explain the advantages and shortcomings of a telemetry patient moni-
toring system.
Explain the two algorithms of ECG arrhythmia detection.
List the desirable characteristics of a patient monitoring network.
Differentiate ring, bus, and star network topologies.
Differentiate host-terminal, client-server, and peer-to-peer networks.
Describe the characteristics of Ethernet and Token Ring network pro-
tocols.
State the characteristics of different types of network connections.
List the functions of a network interface card, repeater, bridge, and
router.

CHAPTER CONTENTS

1. Introduction
2. Functions of Physiological Monitors
3. Methods of Monitoring
4. Monitored Parameters
 Biomedical Device Technology: Principles and Design

5. Characteristics of Patient Monitoring Systems

6. Telemetry
7. Arrhythmia Detection
8. Patient Monitoring Networks

INTRODUCTION

Since the early 1960s, it has been recognized that some patients with
myocardial infarction or those suffering from serious illnesses or recovering
from major surgery benefit from treatment in a specialized intensive care
unit (ICU). In such units, cardiac patients thought to be susceptible to life-
threatening arrhythmia could have their cardiovascular function continuous-
ly monitored and interpreted by specially trained clinicians.
Technological advances led to the ability to monitor other physiological
parameters. Temperature transducers such as thermistors made it possible to
continuously record patient temperature. Through impedance plethysmog-
raphy, respiration could be monitored using ECG electrodes. The develop-
ment of accurate, sensitive pressure transducers gave clinicians the ability to
continuously monitor venous and arterial blood pressures; improved ca-
theters made it possible to monitor intracardiac pressures and provide rela-
tively easy and safe methods of measuring cardiac output. Noninvasive
means were developed to monitor parameters such as oxygen saturation
level in blood.
As clinical knowledge continues to advance and become more sophisti-
cated, special cardiac care units (CCUs) evolved to centralize cardiovascular
monitoring. When patients in danger of cardiac arrest are grouped together
with trained staff, resuscitation equipment, vigilance, and combined with
prompt responses to cardiac emergencies, lives can be saved. The concept of
intensive specialized care assisted by continuous electronic monitoring of
physiological parameters has been applied to specialties other than cardiolo-
gy, resulting in the formation of other special care units such as pulmonary
ICUs, neonatal ICUs, trauma ICUs, and burn units.

FUNCTIONS OF PHYSIOLOGICAL MONITORS

Patient monitors continuously or at prescribed intervals measure physio-

logical parameters and store them for later review, and thus can free up some
clinicians' time from performing other tasks. In addition, monitors with built-
in diagnostic capabilities can alert clinicians when abnormalities such as
 Physiological Monitoring System

tachycardia (abnormally high heart rate) occur. Most physiological monitors

carry out the following basic functions:
Sense-pick up and transform the physiological signal into a more
machine friendly format (e.g., a pressure transducer to convert blood
pressure waveform to an electrical signal)
Condition-amplify, filter, level shift, etc.
Analyze-make measurements and interpretations of the signal (e.g.,
extract heart rates from ECG waveform)
Display-show the output in visual (e.g., waveforms, numerical values)
or audio formats
Alarm-provide visual and audio warning when some limits are
exceeded
Record-store information on paper or electronic media
In a modern patient care ward such as an ICU, several physiological
parameters of a patient are usually being monitored. Instead of having mul-
tiple single-parameter monitors clustered around a patient, a multiparameter
monitor is used. A multiparameter monitor has the capability to capture sev-
eral physiological parameters simultaneously. In addition to saving space, a
multiparameter monitor is more economical as the modalities may share
some common fundamental components. Figure 14-1 shows a three-para-
meter monitor capable of measuring ECG, blood pressure, and temperature.
Each of the three modules captures (senses and conditions) a physiolog-
ical signal from the patient. As the modules are sharing the remainder of the

User
input Memory
Visual
display
I--------;
I
I
I
ECG
module 1 * b Central
- Audio
L - - - - - - - -1 Z processor alarm
r-------- m.
I Blood n_
tD
-
X
pressure ? A
I module b Paper
.- ----- - -1
Power
chart
recorder
I--------;
supplies
I Temperature b
I module I

.- - - - - - -
I
-1
I
interface

Figure 14-1. A multiparameter Patient Monitor.

 Biomedical Device Technology: Princ$Zes and Design

device components, the signals from the modules are multiplexed and sent
to the central processor. The processor analyzes the signals and extracts nec-
essary information from them (such as heart rate from the ECG, systolic and
diastolic values from the blood pressure waveform). Such information is then
compared with preset parameters to trigger audio or visual alarms. The phys-
iological waveforms as well as numerical information are displayed in the
video monitor and hard copies are created by the paper chart recorder. The
monitor may also be connected to a central monitor or to the hospital infor-
mation system through a computer network.

METHODS OF MONITORING

The simplest form of patient monitoring is shown in Figure 14-2a, where

the patient's physiological parameters are displayed at bedside only. This
assumes either a one-to-one nurse-patient ratio or a physical grouping of
patients that permits viewing of the monitor by the nursing staff. Since
alarms must be at the bedside in this type of system, nursing response must
be immediate to prevent unnecessary psychological stress accompanying an
alarm.
The need for patient privacy in the confines of a single room and the
need to economize on nursing staff led to the use of both bedside and a
remote monitoring station (commonly refers to central monitoring station).
Extending the patient-nurse ratio from 1:1 to 2: 1 or even 3: 1 permits a more
economical approach of surveillance (Figure 14-2b). Under such an arrange-
ment, it is also possible to remove alarms from the patient's range of hear-
ing. Furthermore, the central monitoring and physical layout allows nurses
to maintain visual contact with patients and displayed parameters while still
able to engage in other nursing duties in the work area.
The grouping of monitors permits the economical addition of compo-
nents that can be selectively applied to any bed. For example, through suit-
able grouping, a single chart recorder to provide hard copy rhythm strips can
be configured to serve multiple (e.g., four to six) patient beds. Such a central
recorder can be programmed to print either on demand by the user or auto-
matically in the event of an alarm.

MONITORED PARAMETERS

Some examples of physiological parameters being monitored are:

Electrocardiograph-heart rate, arrhythmia, ST segment level
 Physiological Monitoring Systems

Monitor Monitor
Monitor

Central
rnon~tor ,
'\\
\\
4 /'
/'

\\ I ,/'

Figure 14-2. Methods of Monitoring.

Hemodynamics-systolic, diastolic, and mean blood pressure; cardiac

output
Respiration
Temperature
End-tidal carbon dioxide level (EtC02)
Percentage oxygen saturation (%SaOs)
Blood gas (POz, PC02)
While surface ECG, noninvasive blood pressure, and Sp02 are basic
parameters being monitored in most specialty areas, additional monitoring
needs are required in different areas of care. For example, end tidal carbon
dioxide (EtC02) and bispectral index (level of consciousness) monitoring on
patients under general anesthesia are often required in the operating room.

CHARACTERISTICS OF PATIENT MONITORING SYSTEMS

The purpose of a patient monitoring system is to monitor vital physio-

logical parameters so that clinicians ean be alerted to adverse changes in the
patient's condition and provide appropriate interventions. A typical system
often consists of a number of bedside monitors, a few ambulatory monitors,
a central station, and sometimes a telemetry system. In a modern patient
monitoring systems, all of the above are connected via a patient monitoring
network.
 Biomedical Device Technology: Princ@lesand Design

Bedside Monitors
A bedside monitor is positioned beside the patient bed location to
acquire physiological signals from the patient. The monitor is either mount-
ed on the wall or placed on a shelf beside the patient's bed. Catheters and
leads physically connect the transducers or electrodes on the patient to the
input modules of the monitor. Some of the common features of a bedside
monitor are:
Multiple traces display-able to display more than one trace of the
same or different physiological parameters. For example, a four-chan-
nel monitor can be configured to display two channels of cascaded
ECG, one arterial blood pressure, and a %SaOz waveforms.
Alarms-provide visual and/or audio alerts when physiological vari-
ables are outside certain preset values. Usually have silencing feature
and are able to automatically reset into "ready" mode after powered
OFF and ON.
Freeze capability-able to freeze the waveform displaying on the
screen for more detailed analysis.
Trending capability-receive input from any number of slowly chang-
ing physiological variables and plot a continuous record of this vari-
able over a long period of time. For example, plotting the number of
ectopic beats, heart rate, respiration rate, temperature, and blood pres-
sure over time (1, 8, and 24 hour basis).
Recording-able to record a physiological waveform on a printing
device. The printing device may be integrated with or networked to
the beside monitor.
A bedside monitor can be preconfigured or modular. For a preconfig-
ured monitor, all physiological parameters are built in as an integral part of
the monitor at the factory (e.g., a monitor comes with one ECG, one tem-
perature channel, and two blood pressure channels). For a modular bedside
monitor, each physiological parameter is an individual module. A bedside
monitor can be custom-configured by the clinician at the bedside by select-
ing the modules to meet the monitoring needs of the patient. Modules can
be inserted and removed easily by the user. Although the cost of a modular
monitor is usually higher than that of a preconfigured monitor with the same
features, a modular designed monitoring system is more economical and
provides greater flexibility than that of a preconfigured design. For example,
instead of having cardiac output measurement capability in all the monitors
in a 12-bed ICU, three cardiac output modules can be shared among 12
modular monitors as cardiac output measurements are done on an intermit-
tent basis and not on all patients.
 Physiological Monitoring System

Ambulatory Monitors
Very often, a patient staying in a hospital has to be transported from one
patient location to another or to another hospital. For example, a patient in
the Emergency Department may need to be moved to Radiology to have a
CT scan. To facilitate patients who require uninterrupted monitoring, a
smaller, battery-powered monitor that can be brought along with the patient
is necessary. Ambulatory monitors are special monitors that can be trans-
ported with the patient. Some manufacturers have ambulatory monitors that
use the same bedside monitor modules. Such a system can avoid having to
disconnect the patient cables and catheters from the bedside monitor and
reconnect to the ambulatory monitor. To prepare for transport, a user simply
removes the modules from the bedside monitor (while still connected to the
patient) and inserts them into the ambulatory monitor.

Central Station
The location relationship between the patient bed areas and the nursing
work areas should be one where the nurse is never far from hidher patients
when he/she is carrying out hidher routine tasks away from the patient, and
where he/she can maintain visual observation of the patients. Similarly, the
patient, frequently anxious, gains much reassurance by being able to keep
the staff in view and by knowing that they are never far away should he/she
need help. Therefore, a properly designed central station should maintain
two-way visibility between the nurses on duty and each of hidher patients.
In practice, the central station is an extension of the bedside monitor and
provides information from all patients at one location. Typically, one or
more large multitrace central monitors and one or more chart recorders are
located at the central station. By observing the central monitor, all patient
activities can be observed at a glance. In addition, a chart can be printed
automatically or manually from the central recorder at the central station.
The central monitor is usually a large multitrace instrument capable of
displaying several waveform traces at the same time. A basic central moni-
tor has the following capabilities:
Display multiple traces per monitor
Waveform selection and position controls for each of the traces on the
central display
Waveform freeze capability on all traces
Display digital values indicated at the bedside along with alarm limit
settings
Selective trending of parameters
 Biomedical Device Technology: Pfinc$les and Design

It may also include the following capabilities:

Arrhythmia detection and ST segment analysis
Record keeping (e.g., medication and drug interaction)
Connection to the hospital information system (HIS) for patient infor-
mation downloading, retrieval, and billing
The central chart recorder is interfaced with the monitoring alarm sys-
tem to instantly record the signal if an alarm occurs. It is usually a multi-
channel recorder capable of simultaneously printing a number of traces from
the bedside monitors. The recorder can also be used manually to produce a
printout of selected traces from any beside monitors upon demand or, when
so equipped, can automatically provide a printout at predetermined time
intervals.

TELEMETRY

A conventional ECG patient cable confines the patient to the bedside

monitor and limits hidher mobility. Exercise is considered beneficial to
coronary patients; increase in mobility often increases the rate of recovery.
This limitation can be removed by an ECG telemetry system. A telemetry
system removes hardwired connections by replacing it with radio frequency
links. ECG telemetry was developed in the 1950s for stress testing. It now
replaces or supplements hardwired monitors in acute care units. A typical
ECG telemetry system consists of:
A radio transmitter, about the size of a deck of cards, carried by the
patient on a belt or in a pocket
Surface (or skin) electrodes attached to the patient, which, through
lead wires, feed the ECG signal into the transmitter
A bedside or central station receiver that detects the radio wave and
reconstructs the ECG waveform
An ECG telemetry systems can be found in progressive care and coro-
nary (or cardiac) care (CCU) units. It allows continuous monitoring of pa-
tients who require less care and need mobility. Typically, a CCU patient
whose ECG rhythm has satisfactorily stabilized will be put on telemetry
where ECG monitoring continues to detect dangerous arrhythmia. However,
ECG telemetry has the following shortcomings:
Increased system's complexity and cost
ECG artifacts associated with increased mobility
Decreased reliability (loss of signal, range limit, electrical interference)
Delays in locating patients because the freedom afforded by telemetry
encourages wandering
 Physiological Monitoring Systems

In a typical ECG telemetry system, the patient wears a transmitter, which

is connected to three or more skin electrodes that detect the ECG signal
(Figure 14-3). The ECG signal is modulated and transmitted into the free
space. The receiver at the receiving station decodes the signal and displays
the waveform and heart rate on the monitor. Telemetry is also available for
other physiological parameters.

% Patient
Telemetry
transmitter
Telemetry receiver
Figure 14-3. ECG Telemetry System.

ARRHYTHMIA DETECTION

ECG is an important physiological parameter in patient monitoring.

When a patient has a heart problem such as myocardial infarction, the ECG
waveform is different from that of a healthy individual. A computerized
arrhythmia detection system continuously analyzes the ECG waveforms and
attempts to recognize arrhythmias and to alarm on certain ones. Most patient
monitors with arrhythmia detection capability specifically identify and count
different types of arrhythmia. The computer in the monitor measures QRS
complexes, compares beat intervals, follows some algorithm to classify indi-
vidual beats, and recognizes arrhythmias. The computer may use additional
criteria, such as width, prematurity, heart rate, and compensatory pause, to
further classify the type of arrhythmia. In general, there are two types of
algorithms in arrhythmia detection:
1. Waveform feature extraction
2. Template (matching or cross-correlation variety)
Waveform feature extraction measures several QRS characteristics (e.g.,
width, prematurity, height, and area). This is often used in combination with
 Biomedical Device Technology:Princ$les and Design

a compression (data reduction) technique such as AZTEC (Amplitude-Zone-

Time-Epoch-Coding). AZTEC converts the sample ECG signal to a series of
constant-amplitude or sloped line segments for more efficient input to the
computer (see Figure 14-4). The computer then identifies the QRS complex
and makes the required measurements.

ECG Waveform Sampled ECG AZTEC Transformation

Figure 14-4. AZTEC ECG Compression.

These measurements are then compared to criteria stored in the com-

puter system to differentiate between normal and abnormal beats. As the
computer reviews each beat, it groups those with similar features (e.g.,
height, weight). In most systems, the computer then classifies the beats as
normal, paced, premature, et cetera.
In template matching, the computer samples each beat at approximately
16-40 points. These values are then mathematically matched with the values
from previously stored beats or a general set of templates. In template match-
ing, the beat is classified as normal, abnormal, or questionable based on the
number of points on the sampled beat that violate the template boundaries
(see diagram below).

BOUNDARIES OF
TEMPLATE
I :.
:.-- .>Z-

-?$-
.is.
;z<--
%7
+;-:< :a
*:I
69

i;;
NORMAL BEAT VIOLATION OF TEMPLATE
TEMPLATE EQUALS ABNORMAL BEAT

Figure 14-5. Template Matching Arrhythmia Detection Algorithm.

Template cross-correlation algorithm uses a calculated correlation coeffi-

cient (a number from -1 to +1) to mathematically determine how closely the
beat matches one of a set of stored templates. A correlation coefficient
 Physiological Monitoring Systems

greater than or equal to some criterion (e.g., 1 0.9) constitutes a match

between the sampled beat and the template.
With both template algorithms, questionable beats are usually classified
as noise or artifact if similar beats are not detected within a specified time
period (generally under 1 minute) or within a certain number of beats (e.g.,
500). Those that are seen again are classified as abnormal beats.
All algorithms are subject to error, resulting in misclassification of noise
and artifact as arrhythmia and incorrect categorization of arrhythmia. Since
the "normal" QRS complex of a patient may change with time and condi-
tion, the system must "relearn" the patient's normal QRS from time to time
to minimize false alarms.

PATIENT MONITORING NETWORKS

To connect the central and bedside monitors, a "Patient Monitoring

Network" is required. A "network" is a group of computerized devices con-
nected by one or more transmission media for communication or sharing of
resources and data. The transmission links can be wired or wireless.
Resources to be shared in a network can be hardware (such as hard disks,
printers), software, or human resources (such as accessing remote clinical
experts for consultation). Data to be shared can be any information such as
medical images, physiological waveform, or patient information. A network
can be a local area network (LAN),which operates in a small area, or a wide
area network (WAN), which connects a number of LANs over a large geo-
graphic area.
Some desirable characteristics of a patient monitoring network in health
care applications are:
Easily adaptable and configurable to all site-specific requirements (i.e.,
different number of monitored bedside system, distances, etc.)
Easy and fast to move information throughout the network
Allow modifications or changes to system without loss of required
function of the rest of the system
Disconnect and reconnect equipment to system without disruption
Scalable and allow variability of monitoring parameters
Compatible with other network equipment and system (i.e., adheres to
industry standards)

Network Topologies
Components in a network can be physically connected in different ways.
The following diagrams show some possible network topologies (physical
 Biomedical Device Technology: Princ$les and Design

ways of connections) for a patient physiological monitoring network system.

In a ring topology, data passes through each instrument on the way
around the circle. In a bus (or tree) topology, the main data are on pipeline
or bus. Each computer receives the same information. In a star topology, the
central controller connects to other computers like branches radiating out
from the center. It may include satellites.

Bed 2 + Bed 3

Bed I Bed 4

Central Bed
station

Figure 14-6a. Ring Network Topology.

'7 Central # 1

Bed #I Bed #2 Bed #3 Bed#4 Bed #5

Figure 14-6b. Bus Network Topology.

Figure 14-6c. Star Network Topology.

 Physiological Monitoring Systems

Network Protocols
The ARCnet network protocol, developed in the 1970s by Datapoint,
was once a significant industrial standard to handle data link in networking.
However, due to its low transmission rate (2.5 Mbps), it was slowly taken
over by the Ethernet and Token Ring in the 1980s. Today, Ethernet is the
dominant data link protocol used in computer networking, including physi-
ological monitoring. The characteristics of the ethernet and token ring pro-
tocols are:
Ethernet
Ethernet LANs were first developed by Xerox in the 1970s
Adhere to the IEEE 802.3 Standard
10 Mbps (10 Base-2 Thinnet) to 100 Mbps (Cat. 5 UTP fast Ethernet)
to Gbps bandwidth
Access methodology is CSMA/CD (carrier sensed with multiple
access and collision detection)
All stations share the same bus
A station ready to transmit will listen to make sure that the bus is not
in use
Upon a collision (two or more stations were transmitting simultane-
ously), each will wait for a random period of time and then retransmit
Quite efficient for low traffic LANs
BUS-10 Base-2 (Thinnet) or 10 Base-5 (Thick Ethernet)
STAR-10 Base-T (UTP), center of the STAR is a HUB or concentra-
tor
Token Ring
Token Ring LANs were created by IBM and introduced as the IEEE
802.5 Standard
Called a logical ring, physical star
4 or 16 MBps bandwidth
Uses Token-passing access methodology
Guarantees no data collisions and ensures data delivery
Sequential message delivery as opposed to Ethernet's broadcast deliv-
ery
Contention is handled through a TOKEN that circulates past all sta-
tions
Token Ring LANs can be set up in a physical ring or a physical star
The center of the STAR is called a multistation access unit (MAU)
To handle networking and transportation of information, the TCP/IP
(transport control protocol and internet protocol) is by far the most com-
monly used network protocol today to resolve addresses, route information,
and ensure reliable data delivery.
 Biomedical Device Technology: Princ$les and Design

Networks Models
There are three main network models, each of which is characterized by
how it handles traffic and data.
Host-terminal
A host computer connected to dump terminals
The central host handles processing
Terminals provide display and keyboard input

Intelligent client workstations connected to a server computer

Application can be customized and processed at the workstation
Considered as a distributed processing network
The server provides services such as file and printer access to the
workstations
Can have more than one server, e.g., a file server and a printer server
in a network
Peer-to-peer
Two or more computers connected and running the same network
software
Each can do its own processing
Good for small networks to share resources such as printers, storage,
application software, etc.

Network Operating Systems

Most modern commercial network operating systems (NOS) can work
with Ethernet and Token Ring. Examples are the Novel1 Netware, hhcrosoft
Windows NT and 2000 Servers, UNIX, et cetera.

Network Connection Component.

Transmission Links
The network hardware-and software will determine the data transfer
rates between networks and the components within a network. In LANs or
WANs, the cables connecting the components are often one of the major fac-
tors affecting the data transfer rate. The type of connection and the distance
will limit the maximum data transfer rate. Hardwired systems can use twist-
ed copper wires, coaxial cables, shielded cables, or fiber-optic cables. Wire-
less links can use infrared, radio frequency, or microwave for data transmis-
 Physiological Monitoring Systems

sion. For rapid transmission of a large amount of data (e.g., for video confer-
encing), high-speed links are available through telephone or cable compa-
nies. Many organizations have installed such high-speed links as the back-
bone of their WAN. Examples of high-speed links are Integrated Services
Digital Network (ISDN), which has a transmission rate of 64 to 128 Kbps;
T-1 lines with a rate of 1.44 Mbps; and Fiber Distributed Data Interface
(FDDI) with a rate of 100 Mbps.

Network Interconnection Devices

A number of interconnection devices can be found in a computer net-
work system. Some of the common devices are:
Network interface card (NIC)
The interface between the computer and the physical network con-
nection
Responsible for sending and receiving binary signals according to the
network standards
Each NIC has a unique physical address
Hubs
A device to connect several computers
Signals sent to the hub are broadcasted to all ports (where computers
or network devices are connected) of the hub
Switches
A device to connect several computers
Signals sent to the switch are broadcast to selected ports
Repeaters
A device to extend network cabling segments over longer distances
Basic functions are to receive, amplify, and transmit signal
Bridges
An internetworking device to connect small number of LANs togeth-
er
When a bridge receives a packet or a message, it reads the address and
forward the message if it is not local
Reduce traffic congestion because it will not pass local packets to other
LANs
Routers
An internetworking device to direct traffic across multiple LANs or
WANs
 Biomedical Device Technology: Princ$les and Design

Communicate to each other using a routing protocol that includes a

routing table
The routing table stores information about network accessibility and
optimal routing routes

Health Care Network Standards

With the need to share information, LANs and WANs are installed in
hospitals, clinics, and communities. To allow data transfer between terminal
units connecting to these networks, standards are developed to enhance the
connectivity of different medical devices and systems. Examples of standards
for the exchange of patient and clinical data are HL7 (Health Level 7) for
electronic data exchange in health; and DICOM (Digital Imaging and
Communications in Medicine) for vendor-independent digital medical
images transfers between equipment.
DICOM standard sets the basis for interoperability between imaging
devices that claim to support DICOM features. It facilitates equipment from
different manufacturers to communicate to each other. DICOM 3 supports
a subset of the OSI upper level service and is implemented in software.
HL7 is a standard for the exchange, management, and integration of data
that supports clinical patient care and the management, delivery, and evalu-
ation of health care services. Level 7 refers to the highest level of the ISO-
OSI communication model-the application level. This level addresses defi-
nition of the data to be exchanged, the timing of interchange, and the com-
munication of certain errors to the application. It supports such functions as
security checks, participant identification, availability checks, exchange
mechanism negotiations, and, most importantly, data exchange structuring.
 Chapter 15

ELECTROCARDIOGRAPHS

OBJECTIVES

Explain the origin of ECG signal and the relationships between the
waveform and cardiac activities.
Explain projection of the three-dimensional cardiac vector and analyze
the relationships between the ECG leads.
Define 12-lead ECG, the electrode placements, connections, and their
relationships.
Differentiate between diagnostic and monitoring ECG and explain the
effects of changing bandwidth on the display waveform.
Identify and analyze the functional building blocks of an ECG machine.
Study typical specifications of an electrocardiograph.
Evaluate causes of poor ECG signal quality and suggest corrective solu-
tion.

CHAPTER CONTENTS

I. Introduction
2. Origin of the Cardiac Potential
3. The Electrocardiogram
4. ECG Lead Configurations
5. Standard 12-Lead ECG and Vectorcardiogram
6. Fundamental Building Blocks of an Electrocardiograph
7. Typical Specifications of Electrocardiographs
8. ECG Data Storage, Network, and Management
9. Common Problems
 Biomedical Device Technology: Princ$les and Design

INTRODUCTION

The class of medical instrumentation to acquire and analyze physiologi-

cal parameters is called diagnostic devices. This chapter introduces an
important diagnostic medical device to monitor and analyze the heart con-
dition through collecting and evaluating electrical potential generated from
cardiac activities. This medical device is called electrocardiograph, and the
record of the electrical cardiac potential as a function of time is called the
electrocardiogram. The first ECG came into clinical use in the 1920s using
electron vacuum tube amplification, an oscilloscope for display, and a string
galvanometer for recording. ECG has since evolved into a group of highly
sophisticated devices to acquire cardiac potentials, perform diagnostic analy-
sis and interpretation, as well as provide information storage and communi-
cation.

ORIGIN OF THE CARDIAC POTENTIAL

The natural pacemaker of the heart is a small mass of specialized heart

muscle cells called the sinoatrial (SA) node. The SA node generates electri-
cal impulses that travel through specialized conduction pathways in the atri-
um (Figure 15-1). As a result of this electrical activation, the atrial muscle
contracts to pump blood from the atria through the two atrioventricular
valves into the ventricles.
While causing the atrial muscle to contract, this electrical impulse con-
tinues to travel and eventually reaches another specialized group of cells
called the atrioventricular (AV) node. In the AV node, the electrical impulse
is delayed by about 100 ms before it arrives at the Bundle of His and its two
major divisions, the right and left bundle branches. These branches then
break into the Purkinje system, which conducts the electrical impulse to the
inner wall of the ventricles, causing the ventricles to contract and pump
blood from the right ventricle into the lung and from the left ventricle to the
rest of the body. The time delay of the electrical impulse in the AV node
allows blood to be emptied from the atria to the ventricles before the ven-
tricular contraction. This coordinated contraction of the atria and ventricles
maximizes the throughput of the cardiac contraction. Figure 15-2 shows the
time delay of the electrical stimulation reaching different locations of the
heart conduction pathway.
The contraction and relaxation of the heart due to synchronized polar-
ization and depolarization of the cells in the cardiac muscles produce an elec-
tric current that spreads from the heart to all parts of the body. The spread-
 Electrocardiographs

AV node

S A node

lnternodal pathways

Figure 15-1. The Heart's Electrical Conduction Pathways.

Figure 15-2. Time Delay (sec) of Cardiac Conduction.

ing of this current creates differences in potential at various locations on the

body. Figure 15-3a shows a typical action potential plotted against time
obtained from a pair of electrodes placed on a ventricular muscle fiber bun-
dle under normal cardiac activity. It shows rapid depolarization (contraction)
 Biomedical Device Technology: Principles and Design

and then slow repolarization (relaxation) of the muscle fiber. As there are
many fiber bundles contracting and relaxing at slightly different times in a
cardiac cycle, the result of these electrical potential forms a cardiac vector of
changing magnitude moving in three dimensions with time. The potential
difference measured using a pair of electrodes placed on the surface of the
body is the projection of the cardiac vector to the line joining the two elec-
trodes. The waveform obtained by plotting this potential difference between
a pair of electrodes placed on opposite sides of the heart as a function of time
is called the electrocardiogram (or ECG).

THE ELECTROCARDIOGRAM

An ECG obtained from electrodes placed on the surface of the body (or
skin) is called a surface ECG. A typical surface ECG is shown in Figure
15-3b. It consists of a series of waves (P, Q R, S, and T) corresponding to
different phases of the cardiac cycle. Roughly speaking, the P wave corre-
sponds to the contraction of the atria, the QRS complex marks the beginning
of the contraction of the ventricles, and the T wave corresponds to the relax-
ation of the ventricles. In a normal heart, relaxation of the atria occurs at the
same time as the contraction of the ventricles. The voltage variation due to
atrial relaxation is not visible because of the large amplitude of the QRS
complex. The amplitude of the R wave for surface ECG is about 0.4 to 4 mV.
Typical amplitude is 1 mV with a cycle time of 1 second (60 beats per
minute). Figure 15-4 shows the relationship between the surface ECG and
the depolarization of the heart. In a normal cardiac cycle:
The P wave precedes the depolarization of the atria.
The PQ (or PR) interval is a measure of the elapsed time from the
onset of atrial depolarization to the beginning of ventricular depolar-
ization.
The QRS complex marks the start of the depolarization of the ventri-
cles.
The QT interval marks the period of depolarization of the ventricle.
The T wave reflects ventricular repolarization.
Delay due to total interruption or nonresponsiveness of some part of the
pathway causes changes in the ECG. For example, if a large nonconductive
area develops in the wall of the ventricle, the shape or duration of QRS will
be altered. Any marked cardiac abnormality such as problems with the SA
or AV nodes or in the ventricular conduction pathways will be reflected by
changes in amplitude and shape of the ECG waveform. Surface ECG is an
important diagnostic tool for clinicians to gain insight into different abnor-
 Electrocardiographs

Depolarization

-il---Sc--
as

Figure 15-3. (a) Action Potential of a Cardiac Fiber Bundle and (b) Surface ECG

Figure 15-4. Surface ECG and the Cardiac Cycle.

ma1 heart conditions. Examples of some cardiac arrhythmias (abnormal

heart rhythms) revealed in diagnostic ECG are shown in Figure 15-5.
An electrocardiogram can be used to diagnose physiological conditions
of the heart (e.g., to track heart rhythm and heart rate). Figure 15-5b reviews
a premature ventricular contraction caused by an ectopic focus from the ven-
tricles. Figure 15-5c is the most severe consequence of ventricular condition,
which occurs when each muscle fiber within the myocardium contracts and
relaxes at its own pace with no coordination. In ventricular fibrillation, the
heart loses its ability to pump blood into the circulatory system.
ECG is an important diagnostic tool of the heart. An ECG stress test
 a) Normal H m t Rhythm,

c) V&&iculw Fibrillation.
R p 1 5 4 . Nonnd a d &rhythmic EGG.

(ECG taken when the patient is exercising) is an example of a diagnostic

ECG. When a patient's heart rhythm is monitored while staying in a hospi-
tal, it is called monitoring ECG. In general, diagnostic ECG contains more
information than monitoring ECG due to two major factors:
1. The bandwidth of diagnostic ECG (e.g., 0.05 Hz to 120 Hz) is wider
than that of monitoring ECG (e.g., 0.5 Hz to 40 Hz).
2. There are more leads (projection of the cardiac vector) taken simul-
taneously in diagnostic ECG than monitoring ECG.
The effects of machine bandwidth and lead configurations on ECG will
be discussed in more detail later in this chapter.
In a critical care area in a hospital, monitoring of a patient's ECG pro-
vides the following information:
Early warning signs of more major arrhythmias that may follow
Immediate detection of potentially fatal arrhythmia by means of
alarms
Feedback on the effectiveness of a treatment intervention
Correlation between cardiac rhythm and treatment variables
Permanent record of ECG waveform on a routine basis
When an out-patient's ECG must be monitored over an extended peri-
od of time, an ambulatory ECG (or sometimes called Holter ECG) is used.
 Electrocardiographs

An ambulatory ECG can record the patient's heart rhythm continuously dur-
ing normal daily activities, say, 24 hours. During monitoring, the patient
wears a small ECG machine with a built-in magnetic tape recorder or a semi-
conductor memory. ECG is acquired from skin electrodes attached to the
patient and stored in the memory. After the acquisition period, the memory
is downloaded to a reader terminal by a cardiology technologist and the
ECG is read and interpreted by a cardiologist.

ECG LEAD CONFIGURATIONS

In earlier discussion, we learned that the cardiac vector has a varying

magnitude and pointing to different directions with time; also, an ECG is the
potential difference measured against time from the projection of the cardiac
vector into a direction according to the placement of the pair of electrodes.
If ECG electrodes are connected to the right arm (RA), left arm (LA), and
left leg (LL) of the patient, one projection of the cardiac vector can be
obtained by connecting the electrodes attached to the left leg and right arm
to the input terminals of a biopotential amplifier. A different projection of the
same cardiac vector can be measured between the left arm and the right arm
electrodes and similarly another projection between the left leg and right
arm. These projections (or lead vectors) in the patient's frontal plane can be
approximated by an equilateral triangle called the Einthoven's triangle.
Figure 15-6 shows the projections of the cardiac vector at a certain time
instant on the Einthoven's triangle.
The ECG obtained between the limb electrodes LA (+) and RA (-) is
called lead I (Figure 15-7), between LL and RA is called lead 11, and

Cardiac
Projections of
cardiac vector

Figure 15-6. Projection of Cardiac Vector in the Frontal Plane.

 Biomedical Device Technology: Principles and Design

Instrumentation
amplifier

Figure 15-7. ECG Lead I Measurement.

between LL and LA is called lead 111. Figure 15-8 shows the configurations
of these limb leads. Note the polarities of the electrodes.
If the potential is measured across a limb electrode and the average of
two other limb electrodes, the ECG obtained is called an augmented limb
lead. Figure 15-9 shows the connections of the augmented limb leads aVR,

Figure 15-8. ECG Limb Leads.

 Electrocardiograph

aVL, and aVF (note that R stands for right, L stands for left, and F stands for
foot). The average of the limb potentials is obtained by connecting two iden-
tical value resistors to the limb electrodes and then connected to the invert-
ing input of the instrumentation amplifier. The limb leads (I, 11, and 111)and
the augmented limb leads (aVR, aVL, and aVF) together are called the
frontal plane leads.
The frontal plane leads represent the projection of the three-dimension-
al cardiac vector onto the two-dimensional frontal plane. In order to recon-
struct the entire cardiac vector, the cardiac potential projected onto another
plane is required. Figure 15-10 shows the position of the electrode place-
ments on the chest of the patient to obtain the precordial leads (or the chest
leads). The precordial leads represent the projection of the cardiac vector on
the transverse plane of the patient. To measure the precordial leads, poten-
tial of each of the chest electrodes is referenced to the average of the three
limb electrodes (that is why they are sometimes referred to as unipolar leads).
Figure 1\5-11 shows the connections to obtain the chest leads. Note that all
resistors to the limb electrodes are of equal value. Which precordial lead is
being measured depends on the position of the electrode on the chest of the
patient (Figure 15-10). The six frontal plane leads and the six precordial
leads form the standard 12-lead ECG configuration. A summary of the elec-
trode positions for the standard 12-lead ECG is shown in Table 15-1.
Note that altogether nine electrodes (three on the frontal plane and six
on the transverse plane) are necessary to obtain the 12-ECG leads simulta-
neously. In practice, a tenth electrode attached to the patient's right leg is
used either as the reference (grounded) or connected to the right-leg-driven
circuit for common mode noise reduction (see Chapter 11). Figure 15-12
shows the characteristic ECG waveform from a standard 12-lead measure-
ment.

Table 15-1.
Standard 12-Lead ECG Electrode Placement
- -- --

Electrode Phcement
Lead --
Positive Polarity Negative Polarity
-- - -

I left arm (LA) right arm (RA)

II left leg (LL) right arm (RA)
111 left leg (LL) left arm (LA)
aVR right arm (RA) '/L (LA + LL)
aVL left arm (LA) ' / A (RA + LL)

aVF 'left leg (LL) I/L (RA + LA)

V1 through V6 chest positions '/J (LA + RA + LL)

- --
 Biomedical Deuice Technology: Princ$les and Desip

Figure 15-9. ECG Augmented Limb Leads.

Direction of precordial leads on

the transverse plane (top view)
Positions of electrodes on chest wall
Figure 15-10. ECG Precordial Leads.
 Figure 15-1 1. Connections for the Chest Leads.

Figure 15-12. Standad 12-Led ECG Waveform.

STANDARD 12-LEAD ECG AND VECTORCARDIOGRAM

From the definition of the limb leads, lead I is the difference in potential
between the electrodes attached to the left arm and the right arm. That is:
lead I = ELA- ERA,similarly lead I1 = ELL- ERA,and lead I11 = ELL- ELA.
The sum of Lead I and Lead I11 equals to Lead 11:
+
Lead I + Lead I11 =-(EL*- ERA) (ELL- ELA)= ELL- ERA= Lead I1
This result agrees with the vector relationships between lead I, lead 11, and
lead 111 shown in Figure 15-8.
 Biomedical Device Technology: Princ$les and Desijp

For the precordial lead TG, where n = 1 to 6,

Vn = En - ERA+ ELA+ ELL
3
For the augmented limb leads,

Since lead I (or I) = ELG - ERAand lead I1 (or 11)= ELL- ERA,
am=--
I' + I 9 similarly, one can show that
2

Furthermore, augmented leads can be obtained by subtracting the aver-

age potential af the three limb electrodes from one of the limb electrode
potential:
ELL - E w + E3u + E L L = E U3_ (T
E M + U ) 2- ~ ( ~ E"~ -
3
)=+~vF.
Similarly, one can show that

=XaVL,
ELA- ERA+ELA+ELL and
3 3
ERA - ERA+ ELA+ ELL= XaVR.
3 3
Consider the Wilson network shown in Figure 15-13. If the corners of
this triangular resistive network are connected to electrodes on the light arm,
left arm, and the left leg of the patient, K, Vi-, Via and VF- are equal to:
 Electrocardiographs

These terminals on the network can therefore be used as the negative ref-
erence to measure the augmented and precordial ECG leads. The Wilson
network allows using only one electrode connection at each location. It also
avoids the need to remove and reconnect lead wires and electrodes during
ECG measurement. Figure 15-14 shows the connections to obtain lead I,
lead aVR, and a chest lead. Typical resistance values of R and R1 in the net-
work (Figure 15-13) are 10 kil and 15 kil, respectively.

LL
Figure 15-13. Wilson Network.

Figure 15-15 shows the acquisition block (or patient interface module) of
a single channel 12-lead ECG machine. During operation, it uses a multi-
plexer or a number of mechanical switches to select which two input combi-
nations of electrodes are connected to the instrumentation amplifier. Note
that for this machine only one lead can be measured at a time. In a fully dig-
ital machine, the Wilson Resistor Network may be eliminated. The lead sig-
nals from such a digital system are derived mathematically from the electri-
cal potentials from the individual electrodes using the lead relationships
derived previously.
In order to simultaneously measure more than one ECG lead, more than
one instrumentation amplifiers are usually required. Figure 15-14 shows a
three-channel ECG machine measuring lead I, lead aVR, and one chest lead
simultaneously. In general, to measure all 12 leads simultaneously, the elec-
trocardiograph will need to have 12 sets of instrumentation amplifiers as well
as 12 display channels. Some machines are using sampling and time-division
multiplexing techniques.
Other than the standard 12-lead ECG, other lead systems or lead loca-
 Biomedical Device Technology: Princ9les and Design

LA I

V"

From
chest
electrodes

aVR

LL

Figure 15-14. Use of Wilson Network in ECG Measurement.

v-
I
RA - VR-
LA - Wilson
'

Network VL-
LL --+
M
P

VF-
. U

v1
to X

V6

RL

Figure 15-15. A Single Channel 12-Lead ECG Front End.

tions are used in diagnostic electrocardiography. One such commonly used

lead is the esophageal lead, which is obtained by swallowing an electrode
into the esophagus so that the electrode is directly behind and close to the
heart of the patient. The esophageal electrode is often referenced to the aver-
age of the limb leads.
 Another interpretation of the electrical cardiac activity is the vectorcar-
diogram. It was discussed earlier that the cardiac vector changes in both
magnitude and direction (in three dimensions) as the electrical impulse
spreads through the myocardium. A vectorcardiogram depicts these changes
as a function of time during the cardiac cycle. Figure 15-16 show the magni-
tude and direction of the cardiac vector projected onto the frontal plane at
five different time intervals during the QRS complex (i.e., ventricular con-
traction). The vector at ti is zero, which corresponds to the quiescent time
before the ventricle starts to contract. When the current starts to flow toward
the apex of the heart, causing the ventricle to contract, the cardiac vector
starts to grow in magnitude as well as change in direction. The vectors at
time intervals b to t.5 are shown in Figure 15-16. The elliptical figure (or loop)
traced by the cardiac vector during the QRS interval using the quiescent
point as reference is called the QW-vectorcardiogram. A smaller loop,
referred to as the T-vectorcardiogram,is also produced by the T wave. As the
magnitude of the T wave is about 0.2 to 0.3 mV and happens about 0.25 sec-
ond after the QRS, it is a much smaller loop that appears about 0.25 second
after the disappearance of the QRS-vectorcardiogram. A still smaller P-vec-
torcardiogram can be recorded during atrial depolarization. Like the con-
ventional electrocardiogram, the vectorcardiogram can be used in the diag-
nosis of certain heart conditions.

Cardiac I
vectors I
(arrow) i
I
I

Time

Figure 15-16. QRS-Vectorcardiogram.

 Biomedical Device Technology: Puinc$les and Design

FUNDAMENTAL BUILDING BLOCKS

OF AN ELECTROCARDIOGRAPH

Figure 15-17 shows the functional block diagram of a typical electrocar-

diograph. The function of each block is described below.

I
KI Calibration

I I
I I
i Defibrillator Lead
Electrodes protection Preamp -, selector -b Amplifier - j
,
on
Patient
I
I

1 Right leg
+I
I
I
I
drive I I
I I
I !
User .--G
I
--------r------------ I
,--------------J--------------,
I
I
I
I I I
input I
I t v f t
I
I
I
f Recorder Signal Signal
f f Filters +
; or display processor isolation 4- f
I I

Figure 15-17. Functional Block Diagram of an Electrocardiograph.

Defibrillator Protection
As the ECG electrodes are connected to the patient's chest, they will pick
up the high-voltage impulses during cardiac defibrillations. Gas discharge
tubes and silicon diodes are used for defibrillator protection (see Chapter 11,
Figs. 11-17 and 11-18) to prevent the high-voltage defibrillation discharge
from damaging sensitive electronic components.

Lead-Off Detector
When an electrode or lead wire is disconnected, the output of the ECG
may display a flat baseline with noise. This may be misinterpreted as asys-
tole. A lead-off (or lead fault) detector can prevent such misinterpretation. A
simple lead-off detector is shown in Figure 15-18. In this design, a very large
value resistor (>I00 M a ) is connected between the positive power supply
and a lead wire to allow a small D C current to flow via the electrode through
the patient to ground. Under normal situation, due to the relatively small
 Electrocardiographs

electrode/skin impedance, the DC voltages created at the input terminals of

the operation amplifier are very small and almost of equal value. However,
if an electrode or a lead wire comes off from the patient, the amplifier will be
saturated since the voltage at one input of the amplifier will rise to the level
of the power supply. In this case, the lead-off LED will turn on to alert the
user to a lead fault.

Figure 15-18. Lead-Off Detector.

Preamplifier
The magnitude of surface ECG is from 0.1 to 4 mV. A system, especial-
ly one with long unshielded lead wires, may pick up noise of up to several
mV through electromagnetic coupling. Therefore, it is important to amplify
this small signal as close to the source as possible before it is corrupted by
noise. Most ECG machines amplify the potential signals picked up by the
electrodes in a preamp module or patient interface module located near the
patient.

Lead Selector
The lead selector selects the ECG lead to be displayed or recorded. In a
multichannel machine, the lead selector also configures the sequence and
format of the display or printout.
 Biomedical Device Technology: Princ$les and Design

Amplifier
Typically the magnitude of the ECG at the surface of the body is about
1 mV, but this value may vary substantially from patient to patient. For ex-
ample, the ECG of a critically ill patient may be as low as 0.1 mV or as high
as 3 mV. The electrocardiography must have some means of controlling the
size of the ECG waveform. This is also called SIZE, GAIN, or SENSITIV-
ITY adjustment. Typical sensitivity settings are 5, 10, or 20 mm/mV.

Right-Leg-Driven Circuit
Electrical equipment and wiring near the electrocardiograph may induce
common mode signal of several mV magnitude on the patient's body. The
right-leg-driven circuit is to suppress this common mode signal so that it will
not mask the ECG signal (see Chapter 11).

Calibration Pulse
A built-in reference voltage of 1 mV is applied to the input of the elec-
trocardiograph. This reference signal is displayed on the screen and on the
printout to inform the user that the machine is functioning properly and that
it has the necessary gain to display the ECG signal coming from the patient.

Signal Isolation
The function of the signal isolation circuit is to reduce the leakage cur-
rent to and from the patient through the electrode/lead connection for
microshock prevention. A module consisting of a FM modulator, an opto-
isolator, and a demodulator is commonly used to serve this purpose.

Filter
The frequency bandwidth for a diagnostic quality ECG is from 0.05 to
150 Hz. Such diagnostic mode bandwidth allows accurate presentation of the
electrical activities of the patient's heart. Monitoring mode is used where a
gross observation of the electrical activity of the patient's heart is necessary
but requires little analysis or details. Interference and baseline drift can be
reduced by a bandwidth less than that required for a diagnostic-quality ECG.
For monitoring, a bandwidth of 1 to 40 Hz is reasonable and will allow
recognition of common arrhythmias, while providing reasonable rejection of
artifacts and power frequency (60 or 50 Hz) interference. However, some
 Electrocardiographs

distortion of the ECG will occur. Most electrocardiographs have built-in

upper and lower cutoff frequency selection to allow the user to choose the
optimal bandwidth for the situation. A power frequency rejection filter
(notch filter) can also be switched on or off by the user to minimize power
frequency interference.

Signal Processor
Signal processing functions in ECG machines can range from simple
heart rate detection to sophisticated arrhythmia analysis and classification.
Some common features for signal processing are:
Heart rate detection and alarm
Pacemaker pulse detection
Waveform measurement: PR interval, QRS duration, etc.
Arrhythmia analysis and classification: e.g., occurrence and frequency
of PVC
Diagnosis and interpretation

Recorder or Display
The acquired waveform of diagnostic ECG can be displayed on a moni-
tor (LCD or CRT) or printed out from a paper chart recorder. In either case,
the speed of the waveform traveling across the screen of the monitor or the
speed of the paper in the chart recorder can be adjusted. Typical speeds are
12.5,25, and 50 mm/s. For a multichannel ECG machine, the display format
can be selected to display a combination of ECG leads. For example, a "3 X
4 + 3R" print format from a six-channel paper chart recorder is shown in
Figure 15-19. In this format, the 12 ECG leads are displayed in three rows
of 4 ECG leads. Each of the leads is displayed for 2.5 seconds. In addition,
three leads selected by the user are displayed for the entire 10 seconds.

TYPICAL SPECIFICATIONS OF ELECl?ROCARDIOGRAPHS

The specifications of a typical 12-lead electrocardiograph are:

Input channels: simultaneous acquisition of up to 12 ECG leads
Frequency response: -3dB @ 0.01 to 105 Hz
CMRR. >I10 dB
Input impedance: >50 MLR
A/D conversion: 12 bits
 Biomedical Device Technalogy: Princ$les and Design

Zk./r Urn. Unknown

lm-/mV 76yr 69tn lE3lb M I W L SIWUS RHYTHI U l l W DCWStCUnL PnEn~nmE~ R I C U L U cuw~exes
IWHs Sex I4 @me.' C-uo INPERIOU I N P U M . ACB U D C T E W I l W D
P b Wlxl roo. I noom Wz AYTBROLATERIL I N P W . ACE UHDeTeRtlINeD
"68 A W R U k L BtT
Vvnt. rere
PR Interval
MS d*t..bO"
OTIOls
112
JIWUNI s
.
P-R-T axen 25 -12 263
82
168
..
BPM
as

Rsferrsd by. DR DUIL Uncenl l rm.4

I i
1
1 I .

!
I
I
/
1 I
1 1

1 , / .

Figure 15-19. A "3 x 4 + 3R" Printout of a 12-Lead ECG.

Sampling rate: 2,000 samples/sec per channel

Writer type: thermal digital dot array with 200 dots per inch vertical
resolution
Writer speed: 1, 5, 25, and 50 rnm/sec, user selectable
Sensitivities: 2.5, 5, 10, and 20 mm/mV, user selectable
Printout formats: 3, 4, 5, 6, and 12 channels, user selectable channel,
and lead configurations
Dimensions: 200 (H) X 40 (W) X 76 (D) cm
Power requirements: 90 VAC to 260 VAC, 50 or 60 Hz
Certifications: IEC 601

ECG DATA STORAGE, NETWORK, AND MANAGEMENT

With the advancement in electronic data storage and computer network

technologies, modern ECG machines are capable of electronically stored
and shared information through computer networks. In a hospital, wireless
ECG telemetry, diagnostic review stations, ECG machines, and electronic
 Electrocardiographs

storage can be integrated into an "ECG data management system" via a local
area network (LAN). Multiple hospitals, through wide area network (WAN),
can also be configured to communicate and share resources such as mass
storage or archive. In a paperless cardiology, ECG data can be readily
stored, retrieved, transferred, and viewed at any designated location.

COMMON PROBLEMS

Abnormality in ECG waveform may be grouped into the following three

categories:
1. Artifacts due to electrode problems may be caused by
Improper positioning of electrodes on the patient
Loose contact between the electrode and the patient
Dried-out electrode gel
Bad connection between the lead wire and the electrode
Failure to properly prepare (clean, shave, and abrade) the patient site
for electrode attachment
2. Artifacts due to physiological interference may be caused by
Skeletal muscle contraction
Breathing action
Patient movement
Involuntary muscle contraction (e.g., tremor,
3. Artifacts due to external interference may be caused by
Power frequency interference coupled to the lead wires or as common
mode voltage on the patient (60 Hz interference in North America)
Radiated electromagnetic interference from other equipment (e.g.,
500 kHz interference from an electrosurgical unit)
Conductive interference from the power line or ground conductor
(e.g high-frequency noise from switching power supplies)
Interfering signals from other equipment connected to the patient
(e.g., pacemaker or neural stimulator pulses)
Power interruption and supply voltage fluctuation
Figure 15-20 shows some common artifacts in ECG acquisitions. Figure
15-20a shows a typical ECG waveform with power frequency interference.
One can see 60 even, regular spikes in a 1-second interval if the timescale is
expanded. Severe 60 Hz interference is often caused by improper patient or
equipment grounding. It may also occur when the ECG lead wires are
placed too close to power cables or improperly grounded electrical equip-
 Biomedical Device Technology: Princ$les and Design

a) Power Frequency (60 Hz) Interference.

b) Baseline Wander.

c) Low Amplitude.

d) Muscle Contraction.
Figure 15-20. Common ECG Artifacts.

ment. Turning on the built-in 60 Hz notch filter (if available) can eliminate
such interference. Grouping the lead wires may reduce the interference
amplitude. Figure 15-20b shows an ECG with wandering baseline. This can
 Electrocardiographs

be caused by poor skin preparation, bad electrode contact, dried-out or

expired electrode, patient movement, or patient's respiratory action. Figure
15-20c shows an ECG waveform with abnormally small amplitude. Poor
skin contact, improperly prepared skin, or dried-out electrode may be the
awe. Figure 15-2Qdshows EGG artifacts due ts skeletal muscle contraction.
Muscle artifacts will usually disappear when the patient is relaxed and
calmed down.
 Chapter 16

ELECTROENCEPHALOGRAPHS

OBJECTIVES

Explain electroneurophysiology and the sources of signals.

Outline the signal characteristics and clinical applications of EEG.
Describe different types of EEG electrodes and their applications.
Explain the 10-20 electrode placements and montages.
Identify the characteristics of EEG waveform.
Sketch and explain the functional block diagram of an EEG machine.
Identify causes of poor EEG signal quality and suggest corrective solu-
tions.

CHAPTER CONTENTS

1. Introduction
2. Anatomy of the Brain
3. Applications of EEG
4. Challenges in EEG Acquisition
5. EEG Electrodes and Placement
6. EEG Waveform Characteristics
7. Functional Building Blocks of EEG Machines
8. Errors and Problems in EEG Recording
 Electroencephalographs

INTRODUCrION

Electroneurophysiology refers to the study of electrical signals from the

central and peripheral nervous systems for functional analysis and diagnosis.
These signals are recorded using extremely sensitive instruments to pick up
tiny electrical signals produced by the system. Electroencephalography was
developed by the German psychiatrist Hans Berger in the mid 1920s,
evolved in the 1930s for clinical use, and expanded quickly in the 1940s. It
includes the field of electrocorticography-multichannel recordings from the
exposed brain cortex. Today, there are four main areas in electroneurophys-
iology: electroencephalography (EEG), evoked potential (EP) studies,
polysomnography (PSG),and electromyography (EMG).
Electroencepholography is a procedure in which small electrical signals
produced by the brain are recorded. These signals are generated by the
inhibitory and excitatory postsynaptic potentials of the cortical nerve cells.
An electroencephalograph is a machine that captures these brain signals.
The electrical potential measured due to these signals plotted against time is
called an electroencephalogram (EEG). EEG may be used to diagnose con-
ditions such as epilepsy and cerebral tumors. In EEG measurements, elec-
trodes are generally placed on the skull of the patient, although some clini-
cians may use electrodes that penetrate the skin surface or electrodes that are
placed directly on the surface of the cerebral cortex. EEG signals measured
directly at the surface of the brain or by needle that penetrated the brain are
typically of amplitude from 10 pV to 3 mV, whereas signals acquired on the
surface of the skull are typically from 1 to 500 pV. Figure 16-1 shows a gen-
eral configuration of EEG recording.
Evoked potentials (EP) are performed to analyze the various nerve con-
duction pathways in the body. Stimulation such as a sound or flashing light
is imposed on the subject to initiate a nerve signal transmission. If the signal
is not getting through, a lesion in the particular nerve pathway may be pre-

Electrodes

I Amplifier I-\ Recorder 1

Figure 16-1. General Configuration of EEG Recording.

 Biomedical Device Technology: Principles and Design

sent. EP studies, for example, are useful in diagnosing problems in the visu-
al and auditory pathways.
Polysomnography (PSG) is the study of sleep disorders by recording
EEG, physiological parameters, and various muscle movements. PSG can be
used in diagnosing and treating sleep disorders such as insomnia and sleep
apnea.
Electromyography (EMG) is the study of the electrical activities of mus-
cles and their peripheral nerves. It may be used to determine whether the
muscles are functioning properly or if the nerve conduction pathway is
healthy.
This chapter focuses on EEG, EMG and EP studies are discussed in
Chapter 17.

ANATOMY OF THE BRAIN

The brain is the enlarged portion and also the major part of the central
nervous system (CNS), protected by three protective membranes (the
meninges) and enclosed in the cranial cavity of the skull. The brain and
spinal cord are bathed in a special extracellular fluid called cerebral spinal
fluid (CSF). The CNS consists of ascending sensory nerve tracts carrying
information to the brain from different sensory transducers throughout the
body. Information such as temperature, pain, fine touch, pressure, et cetera
is picked up by these sensors, and delivered via the nerve tracts to be
processed in the brain. The CNS also consists of descending motor nerve
tracts, originating from the cerebrum and cerebellum, and terminating on
motor neurons in the ventral horn of the spinal column.
The three main parts of the brain are the cerebrum, the brainstem, and
the cerebellum. The cerebrum consists of the right and left cerebral hemi-
spheres, controlling the opposite side of the body. The surface layer of the
hemisphere is called the cortex and is marked by ridges (gyri) and valleys
(sulci);deeper sulci are known as fissures. The cortex receives sensory infor-
mation from the skin, eyes, ears, et cetera. The outer layer of the cerebrum,
approximately 1.5-4.0 mm thick, is called the cerebral cortex. The layers
beneath consist of axon and collections of cell bodies, which are called
nuclei. The cerebrum is divided by the lateral fissure, central fissure (or cen-
tral sulcus), and other landmarks into the temporal lobe (responsible for
hearing), the occipital lobe (responsible for vision), the parietal lobe (con-
taining the somatosensory cortex responsible for general sense receptors),
and the frontal lobe (containing the primary motor and premotor cortex
responsible for motor control).
 Electroencephalograph

The brainstem is an extension of the spinal cord, which serves three pur-
poses:
1. Connecting link between the cerebral cortex, the spinal cord, and the
cerebellum
2. Integration center for several visceral functions (e.g., heart and respi-
ratory rates)
3. Integration center for various motor reflexes
The cerebellum receives information from the spinal cord regarding the
position of the trunk and limbs in space, compares this with information
received from the cortex, and sends out information to the spinal motor neu-
rons

APPLICATIONS OF EEG

EEG can be used as a clinical tool in diagnosing sleep disorders, epilep-

sy, and multiple sclerosis. It can also be used in bispectral index monitoring
as an indicator of the depth of anesthesia in surgery. It is an effective tool in
ascertaining a patient's recovery from brain damage or to confirm brain
damage.
A normal EEG usually consists of a range of possible waveforms from
low-frequency, nearly periodic waves with large delta components in deep
sleep, to high-frequency, noncoherent beta waves measured on the frontal
lobe channels during vigorous mental activity. Under a relaxed state, the
EEG is characterized by alpha waves from the occipital lobe channels.
Opening and closing the eyes results in an evoked response. Examples of the
characteristics of some common EEG applications follow.

Brain Death
Absence of EEG signals is a definition of clinical brain death.

Epilepsy and Partial Epilepsy

Epilepsy may be classified into the following categories:
Generalized epilepsy-affects the entire brain
Grand ma1 seizures-large electrical discharges from entire brain that
last from a few seconds to several minutes. It is apparent on all EEG
channels and may also be accompanied by skeletal muscle twitches
and jerks
 Biomedical Device Technology:Principles and Design

Petit ma1 seizures-this is a less severe form of epilepsy with strong

delta waves that last from 1 to 20 seconds in only part of the brain, and
therefore appears in only a few EEG channels

Diagnosing Sleep Disorders

In normal sleep, the alpha rhythms are replaced by slower, larger delta
waves. EEG monitoring can also determine if and when a subject is dream-
ing due to the presence of rapid, low-voltage interruptions, indicating para-
doxical sleep, or rapid-eye-movement (REM) sleep.

CHALLENGES IN EEG ACQUISITION

Measurement of EEG signals using surface electrodes in general is more

difficult than measuring ECG signals because:
The electrical potentials are conducted through a number of nonho-
mogeneous media before reaching the scalp. Table 16-1 lists the val-
ues of resistivity of different body tissues.
Since tissues have higher resistivity than the cerebral spinal fluid (CSF)
that overlies the brain, the CSF is acting as a region of high conduc-
tivity, having a shunting effect on electrical currents.
Muscles over the temporal region and above the base of the skull pro-
vide pathways of high conductivity, allowing the shunting of local volt-
ages well beneath the skin.
Because of this spatial conductivity arrangement, the electrical poten-
tial difference measured actually shows the resultant field potential at
a boundary of a large conducting medium surrounding an array of
active elements (i.e., activities of the nuclei and some axons).
In addition, utilization of any nervous functions would cause or inhib-
it electrical activities of related parts of the brain, leading to a change

Table 16-1.
Tissue's Electrical Resistivity.
Body Part Resistiviv (0-m)
Blood 100-150
Heart muscle 300
Thoracic wall 400
Lung 1,500
Dry skin 6,800,000
- - -
 Electroencephalographs

in the electrical potentials on the scalp. Electrical activities will also be

a function of age, state of consciousness, disease, drugs, and whether
external stimulation is used.
EEG is then, in general, the scalp surface measurements of the total
effects of all the electrical activities in the brain. It is assessed in conjunction
with patient records, clinical symptoms, and other physiological parameters.

EEG ELECTRODES AND PLACEMENT

m e s of EEG Electrodes
Depending on the nature of EEG studies, different types of electrodes are
used. Surface electrodes, due to their noninvasive application, are the most
commonly used electrodes. Needle, cortical, subdural, and depth electrodes
are examples of invasive electrodes. Common materials for EEG electrodes
are Ag/AgCl, Au-plated Ag, stainless steel, and platinum. The constructions
and placements of some are described next.
In an EEG measurement using surface (or scalp) electrodes, the elec-
trodes are made to be in contact with the scalp of the patient. Electrodes may
be in the form of a flat disk (1 to 3 mm in diameter) or a small cup with a
hole at the center for injection of electrolyte gel. Materials such as platinum,
gold, silver, or silvedsilver chloride are used for EEG surface electrodes.
Earlobe electrodes and nasopharyngeal electrodes are some of the other
noninvasive electrodes. In order to minimize noise and artifact problems,
surface electrodes must be affixed to the scalp. One of two methods can be
used:
1. Using collodion (a viscous fluid) to attach the electrode to the scalp.
It is applied to the electrode site and dried using a jet of air.
Electrolyte gel is then injected into the electrode through a hole in
the center. Low-melting-point paraffin may be used as a substitute for
collodion.
2. Adhesive conductive paste is placed directly on the desired location
with the electrode pressed into the center of the paste.
Needle electrodes are sharp wires usually made of steel or platinum.
There are inserted into the capillary bed between the skin and the skull bone.
They can be applied quickly and provide slightly better signal quality than
scalp electrodes. Although it is relatively safe, EEG measurement using nee-
dle electrodes is an invasive procedure.
Cortical electrodes are used during neurosurgical procedures such as
excision of epileptogenic foci. They are applied directly onto the surface of
 Biomedical Deuice Technology: Princ$les and Design

the exposed cortex. These electrodes consist of metal balls or saline-soaked

wicks that may be held in place by swivel joints mounted on a bracket of a
head frame for easy three-dimensional positioning. Figure 16-2 shows an
example of the setup.

Electrode

Figure 16-2. Cortical Electrodes.

Subdural electrodes are used to localize epileptiform activity and to map

cortical function. They consist of a number of disk electrodes mounted on a
thin sheet of flexible translucent SilasticTMrubber. The electrodes are made
of platinum or stainless steel. Subdural electrodes are often configured as lin-
ear strips or rectangular grids with a number of electrode contact points.

Lead wires Electrodes

Figure 16-3. Subdural Electrodes (a) Grid, (b) Single Column Strip.
 Electroencephalographs

They are designed to be placed on the surface of the cortex. A single column
strip can also be inserted into the intracranial cavity through a small burr
hole opening. Subdural grids are placed over the cortical convexity in open
cranial procedures to cover a large surface area. Figure 16-3 shows such elec-
trodes.
Depth electrodes are fine, flexible plastic electrodes attached to wires
that carry currents from deep and superficial brain structures. These currents
are recorded through contact points mounted on the walls of the electrodes.
Fine wires extending through the bores of the electrodes are inserted with
stylets placed in the bores. Stereotactic depth electrodes are useful, for exam-
ple, in determining the site of origin in temporal lobe epilepsy and as stimu-
lating electrodes for the treatment of movement disorders. Either local or
general anesthesia is applied when the electrodes are being inserted into the
brain.

Wires to connector

Conductive contact points

Figure 16-4. Depth Electrodes.

Risks of Implanted EEG Electrodes

Infection is considered the major risk of implanted EEG electrodes.
Recent studies reported an infection rate of about 2-3010. Meticulous surgical
techniques and procedures to prevent cerebrospinal fluid leakage keep the
risk of infection low.
Another risk from electrode placement is hemorrhage. Significant intrac-
erebral hemorrhages have been reported, but the incidence is 10/o or less.
Direct brain injury due to passing of the depth electrodes has not been
demonstrated because the electrodes are so thin that they normally dissect
neural tisue without imposing much injury.

Surface (or Scalp) Electrode Placement

The international 10-20 system of electrode placement provides uniform
coverage of the entire scalp. Based on the proven relationship between a
measured electrode site and the underlying cortical structures and areas,
electrodes are symmetrically spaced on the scalp, using identifiable skull
anatomical landmarks as reference points. It is termed "10-20" because elec-
trodes are spaced either 10% or 20% of the total diitanee between a given
pair of skuIl landmarks. These landmarks are:
Nasion-the root of the nose

- Inion-ossification or bump on the occipital lobe

Right auricular point-right ear
Left auricular point-left ear
F i w 16-5 shows the locations of the electrodes in the 10-29 system
and Table 16-2 lists the names of the electrade positions.

Table 16-2.
- - -
Nomenclature
- -
for
-
the 10-20 System.
- - -- . --
&a& LBadx

----
Brain Arm Lej2 Hemisphere Midline fight Hmriipphen
---
Frontal Pole F ~ l F ~ 2
Frontal F3 F4
Inferior Frontal F7 F8
Mid-Frontal
Mid-Temporal
Posterior T ~ r a t
Central
Vertex or Mid-Central
Parietal
Mid-Parietal
CTcctpital
Non Scalp ha& (~fhsncc)
Auricular A1 A2
Nasopharyngeal* Pg1 Pg2
~~~~~

Note: * optional leads

 Electroencephalograph

lnion Top of Head

Vertex
_---
20%
,/- cz
-.
,
\
1, 20%
0 I

Left Side of Head

Nasopharyngeal lead

Figure 16-5. International 10-20 System Electrode Placement.

 Biomedical Device Technology: A.inct$les and Design

The use of the 10-20 system ensures reproducible electrode placement

to allow more reliable comparison of EEGs from the same patient or differ-
ent patients. Additional electrodes may be placed between a pair of adjacent
electrodes to more accurately localize an event or abnormality.

Scalp-Electrode Impedance
As the EEG signal is of such low amplitude, the impedance of each elec-
trode should be measured before every EEG recording. The impedance of
each electrode should be between 100 IR and 5 kIR. Impedance below 100 IR
indicates short circuit set up by conductive gel between two electrodes.
Impedance above 5 kIR signals poor electrode skin contact. In practice, elec-
trode impedance is usually measured using an ohmmeter by passing a small
alternating current from one electrode through the scalp to all other con-
nected electrodes. An alternating current of approximately 10 Hz is used to
avoid electrode polarization and prevent measurement error due to DC off-
set potential. If only one pair electrodes are used, the impedance should be
between 200 IR and 10 k a . In addition, minimizing the differences in imped-
ance at the different electrode sites can reduce EEG signal size variations.

EEG WAVEFORM CHARACIXRISTICS

The peak to peak amplitude of EEG waveforms measured using scalp

electrodes lies between 0 and 500 pV.A wide variety of EEG patterns can
be seen in different persons of the same age. An even greater variation can
be found in different age groups. An EEG is considered normal if there is no
abnormal pattern known to be associated with clinical disorders. An EEG
with no abnormal pattern does not guarantee the absence of problems as not
all abnormalities of the brain produce abnormal EEG. Figure 16-6 shows an
EEG recording with normal rhythm followed by a run of epileptic events.
The frequency of EEG waveforms can be divided into four frequency
ranges-beta, alpha, theta, and delta. The frequency bandwidths, locations,
and conditions of acquisitions of these four bands are listed in Table 16-3.

FUNCTIONAL BUILDING BLOCKS OF EEG MACHINES

A very basic single channel electroencephalograph is shown in Figure

16-7. In practice, an EEG machine for use in a diagnostic laboratory contains
 Electroencephalograph

I .*x.c '
Figure 16-6. Normal and Abnormal (Epileptic) EEG.

Table 16-3.
Frequency of EEG Waves.
--

Wavefm -
Frequenv
-- -
(Hd
- -
Remarks
- - -
Beta Rhythm 13-30 Frontal-parietal leads
Best when no alpha
Prominent during mental activity
Alpha Rhythm Parietal-occipital
Awake and relaxed subject
Prominent with eyes closed
Disappear completely in sleep
Theta Rhythm Parietal-temporal
Children 2-5 years old
Adults during stress or emotion
Delta Rhythm Normal and deep sleep
Children less than 1 year old
Organic brain &ease

more functions and options. Figure 16-7 shows the functional block diagram
of a typical EEG machine. Their functions are discussed in this section.
 Biomedical Device Technology: Pfinc$les and Desie

d.I""""-, digital
User Input and Control

1
Signal
Filters
lsolatlon

Montagelelectrode
selector
Impedance
tester
s Memory

Electrodes

Subject

Figure 16-7. Functional Building Block of an EEG Machine.

Electrode Connections and Head Box

A head box is used to interface electrodes from the skull to the switching
system. Each lead wire from the electrode applied to the skull is plugged into
the corresponding location on the head box. An EEG referential head box
for standard EEG application accepts 23 electrodes plus a few spares. The
head box contains the first level of signal buffering and amplification to
increase the signal level and provide a high input impedance to minimize
common mode noise. Input impedance of a modern EEG machine is on the
order of tens of M a .

Montage and Electrode Selector

Multichannel recordings are used to determine the distribution of elec-
trical potential over the scalp. In order to gain insight into the activity at a
given location, multiple differential signals from different combinations of
electrode pairs are required. A montage consists of a distinct combination of
differential signals of such multiple channel recordings.
Electrodes are attached in groups of 8 (or 10) in a montage. Because of
this, EEG machines usually have 8 or 16 (or 10 or 20) differential amplifiers.
There are two types of amplifier input connections:
 1. Bipolar connection-measurements taken between two electrodes
2. Unipolar connection-all measurements have a common reference
point
Figure 16-8 shows a unipolar connection. A common reference elec-
trode for unipolar connection is the auricular or nasopharyngeal electrode.
To facilitate grouping of electrodes, an electrode selection circuit is available
at the front end of the EEG machine. Figure 16-9 is a diagrammatic repre-
sentation of a multichannel electrode selector matrix. Any two electrodes can
be switched to the input of any of the differential amplifiers. To facilitate
diagnosis, certain electrode combinations are grouped together to form stan-
dard montages. Depending on the design, montage selection can be done
digitally instead of using analog switches. Standard montages are usually
built into EEG electrode selection function and can be programmed or mod-
ified by the user. Figure 16-10 shows the standard "referential" and "trans-
verse bipolar" montages.

,
\
-
L

>
i?
3 -./
cn
0
L db 3
u
n,
2
0
a
- =
(D
L
w
3
(D 4D
cn
-.
,- -
&

>
4b

Reference
Figure 16-8. Unipolar Connection.

Amplifiers
The amplifier increases the signal level to the desired amplitude for the
analog to digital converter and the display. Together with the digital pro-
cessing circuit, it allows the operator to select a different level of sensitivities;
 Biomedical Device Technology: Princ$les and Design

---- a

4 Channel 1
---- A a

'I
4 r t-
----
4 Channel 2
---- a A

f
t rr I
I
I I
I I I
I
! !

v p
I

---- Channel i

To scalp electrodes

Figure 16-9. Electrode Selection Matrix.

most EEG machines have two ranges of sensitivities: mV/cm or pV/mm. A

common sensitivity setting for general applications is 7 pV/mm. Each chan-
nel of an EEG machine consists of a high gain differential amplifier with a
gain of approximately 10,000. Depending on the number of channels, an
EEG machine typically has 8, 10, 16, or 20 differential amplifiers.

Analog to Digital Converter

The ADC samples and converts the analog EEG signals to digital format
so that they can be processed by the digital computer. Typical sampling rates
is 1,000 samples/second with 12 bits (4,096 vertical steps) resolution.

Signal Isolation
The patient-connected parts are isolated from the power ground via opti-
cal isolators. Signal isolation prevents electric shocks (micro- and
macroshocks) by reducing the amount of leakage current flowing to and
from the patient.
 Electroencephlographs

lnion Transverse Bipolar

lnion Referential
Figure 16-10. Transverse Bipolar and Referential Montages.
 Biomedical Device Technology: Principles and Design

Filters
The signal bandwidths are individually selectable through software
(older machines use analog filters). The high pass filter (low filter) is usually
adjustable in steps from 0.1 to 30 Hz and the low pass filter (high filter) from
15 to 100 Hz. In addition, a notch filter (60 Hz in North America and 50 Hz
in Europe and Asia) can be selected to reduce power frequency noise from
line interference.

Sensitivity Control
Sensitivity of each channel can be adjusted individually to match the
input signal amplitude and the output display. Typical sensitivity range is 2
to 150 pV/mm or 2 to 150 mV/cm. A sensitivity equalizer control allows ver-
ification of the accuracy of all channel sensitivities for the same input cali-
bration signal.

Memory, Chart Speed, Display, and Recorder

Chart speeds of 10, 15, 30, and 60 m d s e c are supported in most
machines. A mechanical paper chart recorder with ink styli on 11 X 17 inch
fanfolded paper was used in older EEG machines. The huge volume of
paper EEG records generated from each study used to create storage prob-
lems in EEG departments. With digital technology in the newer generation
of machines, the EEG signals are stored in electronic memories and dis-
played on CRT or flat panel displays instead of written on paper. To further
reduce the storage requirement, the neurologist may choose to remove non-
pertinent EEG records and save only waveforms containing useful diagnos-
tic information as patient records. To produce paper copies, laser printers are
commonly used for digital EEG machines.

Electrode Impedance Tester

A small 10 Hz alternating current is used to measure the impedance of
the electrode skin contacts. Electrode impedance testing is available on-
demand in real-time modes on individual or all-channel basis. As discussed
earlier, the impedance of a pair of electrodes should be between 200 fl and
10 kR.
 Electroencephalographs

ERRORS AND PROBLEMS IN EEG RECORDING

EEG Artifacts
In EEG measurements, recorded signals that are noncerebral in origin
are considered as artifacts. Artifacts can be either physiological or nonphys-
iological. Physiological artifacts arise from normal biopotential activities or
movement activities of the patient. The primary sources of nonphysiological
EEG artifacts include external electromagnetic interference and problems
with the recording electrodes. While device hardware malfunction may
cause problems, it is not a common source of EEG artifacts. Common
sources of EEG artifacts are:
Artifacts due to physiological interference may result from:
The heart potential results from either patient touching metal and cre-
ating second ground or pulsatile blood flow in brain
Tongue and facial movement
Eye movement
Skeletal muscle movement (uncooperative patient or fine body
tremors)
Breathing
High scalp impedance
Possible Solution:
Must ensure that patient is calm; try to get him or her to relax
Artifacts due to electrode problems may result from:
Improper electrode positioning
Poor contact causing sharp irregular spikes, or the pickup of 60 Hz
noise
Electrodes not secured properly
* Dried-out electrode gel
Oozing of tissue fluids in needle electrodes
Frayed connections
Sweat resulting in changing skin resistance
Possible Solution:
Replace electrodes on scalp; ensure that electrode impedances are
good (less than 10 kfl between electrode pairs)
Artifacts due to electromagnetic interference (EMI) may result from:
60 Hz common-mode interference
Radio frequency interference due to presence of electrical devices
(e-g., an ESU)
 Biomedical Device Technology: Principles and Design

Defibrillation
Presence of pacemakers and neural stimulators
Possible Solution:
Look for proper grounding (no grounding or multiple ground loops)
as well as shielding
Remove sources of EM1
Perform procedure in special EM1 shielded room

Troubleshooting an EEG Problem

Troubleshooting EEG problems is similar to other troubleshooting sce-
narios having surface electrodes. Some common areas to look into are:
Common-mode problems
Problems compounded due to small signal levels (thousand times
smaller than ECG)
Problems with electrodes and leads (multiple)
Use internal calibration control to check internal electronics and dis-
tinguish them from electrode problems
Isolate problem to one channel, use internal calibration-if output is
healthy, then electronics after calibration must be working properly
For problems that happen with one channel only, can rule out com-
mon components such as power supply
 Chapter 17

ELECTROMYOGRAPHY AND EVOKED

POTENTIAL STUDIES

OBJECI'IVES

Explain EMG and EP studies.

State clinical applications of EMG and EP studies.
Sketch and explain a typical functional block diagram of an EMG/EP
machine.
Describe the constructions and applications of different types of surface
and needle electrodes.
Explain the characteristics of electrical activities from needle electrode
examinations.
Differentiate between motor response and sensory nerve action poten-
tial.
Name common signal procession functions used in EMG/EP studies.
Explain the purpose of signal averaging and how it can reduce noise
level in EP studies.

CHAPTER CONTENTS

1. Introduction
2. Clinical Applications of EMG and EP Studies
3. Electrodes
4. Signal Characteristics
5. Machine Settings
6. Signal Processing
 Biomedical Device Technology: Princ$les and Design

In the previous chapter, the applications, signal acquisition, and func-

tional building blocks of EEG were discussed. This chapter introduces elec-
tromyography (EMG) and evoked potential (EP) studies, which are other
areas in electroneurophysiology. EMG studies the muscles and the nerves
that innervate the muscles; EP studies look at the relationships between
nerve stimulations and their responses.

CLINICAL APPLICATIONS OF EMG AND EP STUDIES

An EMG study may be used to establish the relationships between the

signal morphology to the biomechanical variables. An example is compar-
ing the biopotential signal frequency to muscle tension. In an EP study, a
nerve may be stimulated by an electrical signal at one end and the reaction
measured somewhere along the nerve itself to determine the time-location
relationship between the stimulus and the response. Parameters such as
nerve conduction velocity can also be determined.
There are two main areas of applications of EMG and EP studies: one is
in kinesiology and the other in diagnostic.
In kinesiology, the main areas of interest are:
Functional anatomy
Force development
Reflex connection of muscles
In electrodiagnostic, areas of analysis may involve:
Creation of strength-duration curves to assess nerve and muscle
integrity
Determination of nerve conduction velocity to test for nerve damage
or compression
Analyzing firing characteristics of motor neurons and motor units,
including analysis of motor unit action potentials (MUAPs) to detect
signs of pathology such as fibrillation potentials and positive sharp
waves
Figure 17-1 shows a typical configuration of an EMG/EP study. The
EMG or EP is picked up by a pair of electrodes, one being the sensing elec-
trode and the other acting as the reference. The signal is amplified,
processed, and sent to a chart recorder. Depending on the type of studies, a
stimulation signal may be applied.
 Electromyography and Evoked Potential Studies

Sensing electrode

Figure 17-1. General Configuration of EMG/EP Recordng.

ELECTRODES

A number of different electrodes may be used in EMG/EP measure-

ments. The basic electrodes, including grounding, stimulating, and record-
ing, are described next.

Grounding Electrode
As with all work involving electrical equipment, a ground electrode must
be used. Grounding is essential for obtaining a response that is relatively free
of artifact. In general, the ground electrode should be placed on the same
extremity that is being investigated. The ground electrode in EP studies
should be placed, if possible, halfway between the stimulating electrode and
the active recording electrode. Usually the ground is a metal plate that is
much larger than the recording electrodes and provides a large surface area
of contact with the patient. Some clinicians may use a noninsulated needle
inserted under the patient's skin. One should be careful not to apply more
than one ground to the patient at any time. The presence of multiple grounds
from different electrically powered devices can form "ground loops," which
may create noise in the measurement.

Stimulating Electrode
In most cases, a peripheral nerve can be easily stimulated by applying
the stimulus near the nerve. Therefore, most nerve stimulation is done to seg-
ments of nerve that lie close to the skin surface. Because of the need for prox-
imity, the number of nerves accessible to the stimulation and the locations of
the stimulation of that nerve are limited.
 Biomedical Device Technology:Princ$les and Design

Figure 17-2. Stimulating Electrodes.

The stimulating electrodes are normally two metal or felt pads placed
about 1 to 3 cm. apart (Figure 17-2). The electrodes are placed on the nerve
with the cathode toward the direction in which the nerve is to conduct. The
stimulation amplitude is adjusted until a maximal response is obtained and
then by 25 to 50 percent more to ensure that the response is truly maximal.
One may use a needle electrode to stimulate nerves deep beneath the skin.
Other than electrical stimulation, visual or audible stimulations may be used
in EP studies.

Recording Electrodes
Positioning of recording electrodes depends on the type of response
being studied. In motor response recording, the active electrode is placed
over the belly of the muscle being activated. This placement should be over
the motor point to give an initial clear negative deflection (upward) in the
response. In testing of sensory nerve, the active electrode is placed over the
nerve itself to record the nerve action potential. The reference electrode is
placed distal from the active electrode.
In motor response recording, surface electrodes may be used. Surface
electrodes can be made of pure metal or Ag/AgCl in the shape of a circular
disk of 0.5 to 1 cm in diameter. Surface electrodes such as flat buttons, spring
clips, or rings are frequently used in sensory recording. However, bare-tip
insulated needle electrodes placed close to the nerves are also used by many
investigators. Some of the surface electrodes are shown below.
 Electromyography and Evoked Potential Studies

Figure 17-3. Surface Electrodes.

Needle Electrodes
Needle electrodes are commonly used in EMG/EP studies. They are
used to evaluate individual motor units within a muscle to avoid picking up
signals from other muscle units. The following paragraphs describe a few dif-
ferent types of needle electrodes.
A monopolar needle electrode has a very finely sharpened point and is cov-
ered with TeflonTM or other insulating material over its entire length, except
for a tiny (e.g., 0.5 mm) exposure at the tip. The needle serves as the active
electrode, and a surface electrode placed on the skin close to it serves as a
reference. The main advantage of monopolar needle electrodes is that they
are of small diameter and the TeflonTM covering allows them to easily insert
into and withdraw from the muscle. Moving the needle causes less discom-
fort to the patient. However, repeated use of this electrode changes the size
of the bare tip, thereby limiting the number of examinations for which it can
be used. Because the active electrode tip and the surface electrode are sepa-
rated by some distance, it is easier to pick up background noise from remote
muscle contractions.
A concentric needle electrode consists of a cannula with an insulated wire
inserted down the middle. The active electrode is the small tip of the center
wire, and the reference electrode is the outside cannula. Concentric needles
may have two central wires (bipolar), in which case the active and reference
electrodes are at the tip and the outside cannula acts as the ground. Because
the active and reference electrodes are closer together, only local motor unit

Figure 17-4. Monopolar Needle.

Figure 17-5. Concentric Needle.

 Biomedical Device Technology: Principles and Design

Figure 17-6. Single-Fiber Needle.

action potentials (MUAPs) are picked up by the electrode. Another advan-

tage of this electrode is that no reference surface electrode is needed. The
main disadvantage of the concentric electrode is that it has a larger diameter
relative to other needle electrodes. Large diameter needle electrodes tend to
cause more pain and are uncomfortable to move around.
Single-jber needle electrodes are used for special studies. A single-fiber nee-
dle consists of a thin (e.g., 0.5 mm) stainless steel cannula with a fine (e.g., 25
pm) platinum wire inside its hollow shaft. The cut end of the platinum wire
is exposed from a side opening near its tip.

SIGNAL CHARACI'ERISTICS

Most EMG signals have repetition frequencies in the range of 20 to 200

Hz. A single motor unit action potential (MUAP)has amplitude on the order
of 100 pV and duration of 1 msec. The following paragraphs describe the
electrical activities from needle electrode examinations.

The Muscle at Rest

Insertion Activity
Insertion activity is the response of the muscle fibers to needle electrode
insertion. It consists of a brief series of muscle action potentials in the form
of spikes. It is caused by mechanical stimulation or injury of muscle fibers,
which may disappear immediately or shortly after (a few seconds) stopping
needle movements.

Spontaneous Activity
Any activity beyond insertion constitutes spontaneous activity. It can be
due the normal end plate noise, or to the presence of fasciculation (the ran-
 Electromyography and Evoked Potential Studies

Figure 17-7. Needle Electrode Insertion Activity.

End plate noise Jmpv

10 msldiv

End plate spikes

Figure 17-8. Needle Electrode End Plate Noise.

dom, spontaneous twitching of a group of muscle fibers or a motor unit).

Spontaneous activity due to end plate noise when the needle is in the
vicinity of a motor end plate can be monophasic (end plate noise) or bipha-
sic (end plate spikes) potentials (a phase refers to the part of the wave
between the departure and return to the baseline). The monophasic poten-
tials are of low amplitude and short duration. The biphasic activity consists
of irregular short duration biphasic spikes with amplitude of 100-300 pv.

The Muscle During Voluntary Effort

Voluntary activity can be divided into three stages of effort: mild, mod-
erate, and full. Individual motor units can be studied with mild and moder-
ate voluntary effort, while interference patterns are studied during full vol-
untary effort. Motor unit potentials are best studied with the same filter set-
ting used for insertion and spontaneous activity (i.e., 16-32 Hz low cutoff and
8,000 Hz or more high cutoff frequency).
 Biomedical Device Technology: Princ$les and Design

Figure 17-9. Mild Voluntary Effort.

Figure 17-10. Moderate Voluntary Effort.

Mild Effort
Only a few motor units are observed at this stage. These are the smaller
motor units as they are the ones to be recruited first. Amplitude, duration,
and number of phases of individual motor units are measured.

Moderate Effort
The frequency and recruitment of motor units are best assessed during
this stage. Motor units seen at this stage are larger than those seen with mild
effort. As muscle effort increases, motor unit firing rates are increased and
new motor units are recruited.

Full Effort
At maximum contraction, it is difficult to distinguish individual motor
units as the firing rates are high and many motor units are recruited. When
all the motor units are recruited a complete interference pattern is observed.
 Electromyography and Evoked Potential Studies

Figure 17-1 1. Full Voluntary Effort.

Motor Responses
A motor response is obtained by stimulating a nerve and recording from
a muscle that it innervates. The muscle selected should have a fairly well-
defined motor point and be isolated from other muscles innervated by the
same nerve. The excitation of nearby muscles may alter the response and
make it difficult to determine the exact onset of the desired motor response.
A motor response may be characterized by its amplitude, duration, and wave
form. The amplitude is measured from the baseline to the top of the negative
peak (upward) of the motor response. The latency is measured from the
onset of the stimulus to the point of takeoff from the baseline. In motor
response studies, it is important to ensure maximal motor response by using
supramaximal stimulation of the nerve (i.e., using 15 to 20 percent more than
the minimum level of stimulation). The number and size of muscle fibers

Motor response
Stimulation

( Latency
1
b

i","

2 ms
Figure 17-12. Typical Motor Response.
 Biomedical Device Technology: Princ$les and Design

being activated determine the amplitude of the response. Decrease in the

number of motor units or muscle fibers responding will affect the amplitude.
The usual motor response has a fairly simple waveform. It may have one or
two initial negative (up) peaks (the latter usually indicating two muscle being
stimulated) and usually will be followed by a positive deflection (down)
toward the end. If there is dispersion of the times when the motor units dis-
charge, then the amplitude will be lowered and the response spread in time.
The motor response also changes in relationship to the point of nerve stim-
ulation. The more proximally the nerve is stimulated, the lower the ampli-
tude and the longer the duration of responses.

\I
I
-< Latency >-

Figure 17-13. Typical Nerve Action Potential.

Sensory-Nerve Action Potentials

Sensory-nerve action potentials (NAPS)are obtained by stimulating a

nerve and recording directly from it or one of its branches. The recording
site must be remote from muscles innervated by that same nerve because
muscle responses will obscure the much smaller NAP.
A NAP is characterized by its amplitude, duration, and wave form. The
amplitude of the NAP is measured from the peak of the positive deflection
to the peak of the negative deflection. The sensory distal latency is tradition-
ally measured from the stimulus artifact to the takeoff or the peak of the neg-
ative deflection. When conduction velocities are calculated, the same takeoff
of the proximal and distal responses should be used to determine the laten-
cy. The amplitude depends on the number of axons being stimulated and the
synchrony with which they transmit their impulses. If the axons transmit
impulses at comparable velocities, the response duration will be short and its
amplitude high. However, if the axonal velocities are widely dispersed, the
NAP duration will be longer and its amplitude lower.
 Electromyography and Evoked Potential Studies

Conduction Velocity
Conduction velocities can be determined by stimulating a nerve at two
points and measuring the distance between those points. This method can
eliminate the neuromuscular transmission time and is used for most motor
nerves. In sensory studies, however, only one stimulation site is normally
used. The conduction velocity (v) can be computed by measuring the dis-
tance (4 in millimeters (mm) between the two recording points and dividing
it by the difference in latency (ms) between the proximal (t,) and distal
recording points (td), as indicated in this equation:

v=- , the result is expressed as meters per second (m/sec.).

td - tp
Conduction velocities are different in different nerves due to different
anatomical conditions. However, several general principles apply to nerve
conduction studies:
The more proximal the segment of nerve being evaluated, the faster
the velocity will be.
If the extremity being tested is cold, the velocity will be slowed and
the amplitude increased. This effect occurs especially in cold weather,
and some provisions for warming the patient and using a fairly con-
stant room temperature should be made.
The shorter the segment between the two stimulation or recording
points, the less reliable the calculated velocities will be, due to a
greater effect on the margin of error by a shorter distance.
Conduction velocities depend mostly on the integrity of the myelin
sheath. In segmental demyelinating diseases, conduction velocities
may drop to below 50°/o of normal values. Axonal loss will slow down
the conduction velocity. Conduction velocity drop due to axonal loss
is usually in the vicinity of 30010 of normal values.

MACHINE SETTINGS

In studying sensory and motor responses, different filter, sweep speed,

and sensitivity settings are used. Sensory studies are performed with the low
frequency setting between 32 and 50 Hz and the high between 1.6 and 3
kHz. The sweep speed is set to 2 ms/div and the sensitivity at 10-20 pV/div.
Motor studies are performed with the low frequency set to 16-32 Hz and the
high frequencies to 8-10 kHz. Depending on the response's latency and
duration, the sweep speed can be set to anywhere between 2 and 5 ms/div
and the sensitivity between 2 and 10 mV/div.
 Biomedical Device Technology: Princ$les and Design

SIGNAL PROCESSING

Signal processing plays an important role in EMG/EP studies. Described

next are some signal processing functions commonly found in EMG/EP
machines.

Filtering
Filters are used to eliminate unwanted signals such as electrical noise and
movement artifact. The frequency spectrum of muscle action potentials lies
between 2 Hz and 10 kHz. In practice, a band pass filter of 20 Hz to 8 kHz
is often used because motion artifacts have frequencies less than 10 Hz and
a high cutoff frequency is necessary to remove high-frequency noise.

Rectification and Integration

Because the raw signal is biphasic or polyphasic, a rectifier is sometime
used to "flip" the signal's negative content across the zero axis, making the
whole signal positive. Integration is used to calculate the area under the
curve for quantization and comparison.

Amplification
A single MUAP has an amplitude of about 100 pV; signals detected by
surface electrodes are in the range of 5 mV; signals detected by indwelling
electrodes are in the range of 10 mV. All these signals must be amplified
before they can be further processed. If a 1 V amplitude signal is required
for the signal processor, an amplifier with a gain of 100 to 10,000 is neces-
sary. Differential amplifiers with high common mode rejection ratio are used
to minimize induced electrical noise, including 60 Hz power frequency
noise, which is within the bandwidth of the signal. In addition, the imped-
ance of the front-end amplifiers must be considerably higher than the imped-
ance of the electrode/skin or electrode/muscle interfaces. Since indwelling
electrodes have very high impedances (due to low surface area), very high
amplifier input impedances (e.g., >100 M a ) are necessary.

Spectral Analysis
Because EMG signal is actually a summation of MUAPs, some close and
some at a distance from the recording electrodes, it is difficult to know which
 Electromyography and Evoked Potential Studies

motor units contribute to the signal. In trying to differentiate normal from

abnormal waveforms, some investigators, using spectral analysis, have tried
to divide the signal's components into its constituent frequencies.

Signal Averaging
Signal averaging is a technique used in EP studies to extract the low
amplitude evoked response from noise. The amplitude of the evoked nerve
response is on the order of pV,while noise can be on the order of mV. This
technique assumes that noise is random and that the evoked responses at the
same location from identical stimulations are the same. Instead of recording
the nerve response from a single stimulus, multiple nerve responses are
recorded from repeating the same stimulation periodically over a period of
time. The response from each stimulus is stored and the average is comput-
ed by an analog or digital computer. As all the nerve responses are the same,
averaging will produce the same response. However, averaging random
noise will reduce or eliminate the noise superimposing on the signal. In prac-
tice, an evoked potential is acquired from averaging 16 or more evoked
responses.
 Chapter 18

INVASIVE BLOOD PRESSURE MONITORS

Explain the origin of blood pressure waveform.

Analyze the relationships between the blood pressure waveform and
the cardiac cycle.
Compare the magnitude and shape of the blood pressure waveform at
different locations in the cardiovascular system.
Describe the clinical setup for invasive blood pressure (IBP) monitor-
ing.
Sketch the block diagram of a typical IBP monitor.
Explain the construction and characteristics of a resistive strain gauge
blood pressure transducer.
Define systolic, diastolic, and mean blood pressure and explain the
principle of detecting these parameters.
Identify common problems and potential sources of error in IBP mea-
surement.

CHAPTER CONTENTS

1. Introduction
2. Origin of Blood Pressure
3. Blood Pressure Waveforms
4. Arterial Blood Pressure Monitoring Setup
5. Functional Building Blocks of an Invasive Blood Pressure Monitor
6. Common Problems and Causes of Errors
 Invasive Blood Pressure Monitors

INTRODUCI'ION

The earliest attempt to record arterial blood pressure was performed in

1773 by Stephen Hales, a British scientist. Hales used an open-ended tube
with one end inserted into an artery in the neck of a horse and the other end
held at a high level (Figure 18-1). The blood from the horse rose to about 8
feet in the tube above the arterial insertion site and fluctuated by 2 to 3 inch-
es between heartbeats. From the experiment, Hales was able to determine
the blood pressure of the horse using the equation P= pgh where p is the den-
sity of the blood, g the acceleration due to gravity, and h the height of the
blood column. This chapter explores the principles and instrumentations of
invasive blood pressure measurements in clinical settings. Although only
blood pressure measurement is discussed, the same principle and, in fact,
similar instrumentations are used in other physiological pressure measure-
ments such as bladder pressure and intracranial pressure.

Figure 18-1. Hales's Experiment to Measure Arterial Blood.

ORIGIN OF BLOOD PRESSURE

In humans, circulation of blood is achieved by the pumping action of the

heart. Atrial contraction pushes the blood from the right atrium through the
tricuspid valve into the right ventricle and from the left atrium through the
mitral valve into the left ventricle. The positive pressure created by the con-
traction of the ventricles forces blood to flow from the left ventricle through
 Biomedical Deuice Technology: Pfinc@lesand Design

the aortic valve into the common aorta and from the right ventricle through
the pulmonary valve into the pulmonary arteries (Figure 18-2). The blood
from the aorta travels through the arteries and eventually reaches the capil-
laries, where oxygen and nutrients are delivered to the tissues and carbon
dioxide and other metabolic wastes are diffused from the cells into the blood.
This deoxygenated blood is collected by the veins and returned to the right
atrium of the heart via the superior and inferior vena cavae. Contraction of
the right atrium followed by the right ventricle delivers the deoxygenated
blood to the lungs. Gaseous exchange takes place in the capillaries covering
the alveoli of the lungs. Carbon dioxide is removed and oxygen is added to
the blood. This oxygenated blood collected flows into the left atrium via the
pulmonary veins and then into the left ventricle to start another round-trip
in the cardiovascular system.

Venous blood , -_ -
To lung

From lung

Aterial blood 4
Figure 18-2. The Heart and Circulatory System.

The heart is the center of the cardiovascular system creating the pump-
ing force. Every contraction of the heart produces an elevated pressure to
push blood flow through the blood vessels. Relaxation of the heart allows
blood to return to the heart chambers. Blood pressure within the cardiovas-
cular system fluctuates in synchrony with the heart rhythm. The maximum
pressure within a cardiac cycle is called systolic blood pressure, while the
lowest is called the diastolic blood pressure. Blood pressure measured in an
 Invasiue Blood Pressure Monitors

artery is called arterial pressure and pressure measured in a vein is called

venous pressure. Although the S.I. unit of pressure is Pascal (Pa), the unit of
blood pressure commonly used is still in millimeter mercury (mrnHg).Figure
18-3 illustrates the timing of the cardiac cycle showing the blood pressures
in the left ventricle, left atrium, and the common aorta. The pressure in the
left ventricle starts to rise when the heart contracts (corresponding to the
QRS complex of the ECG). When the blood pressure in the ventricle is
above the pressure in the common aorta, the aortic valve opens, allowing
blood to flow from the left ventricle into the aorta and then to the arteries.
During the time when the aortic valve is open, the pressure in the common
aorta is virtually the same as that in the ventricle. After the contraction, the
heart relaxes, causing the ventricular pressure to drop rapidly. The pressure
drop in the common aorta is slower than that in the ventricle due to the back-
pressure from downstream and the elasticity of the blood vessels. As the
pressure in the left ventricle falls below the pressure in the common aorta,
the aortic valve (which is a one-way valve) closes, hence separating the left
ventricle from the common aorta. The blood pressure in the common aorta
fluctuates.

Figure 18-3. Events of a Carhac Cycle and Blood Pressure Waveforms.

 Biomedical Device Technology: Principles and Design

BLOOD PRESSURE WAVEFORMS

As the arterial blood flows into smaller blood vessels, the average (mean)
pressure as well as the magnitude of fluctuations (difference between systolic
and diastolic pressure) drop due to friction and viscosity. Arterial blood pres-
sure eventually reaches its lowest level in the capillaries. Venus blood pres-
sure is the lowest just before it enters the right atrium. Figure 18-4 shows the
values of typical mean, systolic, and diastolic blood pressure measured at dif-
ferent locations in the cardiovascular system. Since the left ventricle is the
primary pumping device in the cardiovascular system, the blood pressure is
elevated from the lowest level at the inlet of the left atrium to almost the
highest as it leaves the left ventricle.

140 ,
- . -_..-"
.
120 - t
-- - DtasM~c
M e a n
100 .. --- -Systolic

-r" -
E
E- 80..

f I
22 60.
P I
40- I
,
I
. '.

0,
I L - d
/ - -
'.-

Figure 18-4. Typical Blood Pressure at Different Points of the Cardiovascular System.

Figure 18-5 shows a typical blood pressure waveform. Note that the
blood pressure is referenced to the atmospheric pressure and does not go
negative. Each cycle of fluctuation corresponds to one cardiac cycle. The
characteristic dicrotic notch is a result of the momentum of blood flow and
 Invasive Blood Pressure Monitors

the elasticity of the blood vessels. When the pressure inside the ventricle is
lower than that in the common aorta, the aortic valve closes and suddenly
stops blood from flowing out of the left ventricle. The blood flow continues
for a brief moment right after the valve closure due to the momentum of the
blood velocity. This flow creates a transient pressure reduction in the aorta
as well as in all arteries. The dicrotic notch is less noticeable in smaller arter-
ies and disappears altogether in the capillaries. Within a cardiac cycle, the
blood pressure goes from a minimum to a maximum. The maximum pres-
sure is called the systolic pressure (Ps),while the minimum is the diastolic
pressure (Po).The mean blood pressure (Pdis determined by integrating the
blood pressure waveform over one cycle and dividing the integral by the
period (2).

In some older blood pressure monitors, the mean blood pressure is

approximated by the equation:

Figure 18-6 shows typical blood pressure waveforms at different loca-

tions of the cardiovascular system.

Pressure
(mmHg) +

T Time (sec)
Figure 18-5. Typical Blood Pressure Waveform.
 Biomedical Device Technology: Principles and Design

Arteries

Arterioles

Capillaries

Veins

Figure 18-6. Blood Pressure Waveforms at Different Points

of the Cardiovascular System.

ARTERIAL BLOOD PRESSURE MONITORING SETUP

A typical arterial blood pressure monitoring setup is shown in Figure

18-7. Instead of inserting a pressure transducer into the blood vessel, it
employs a less invasive approach by coupling a liquid column between an
external pressure transducer and the blood in the vessel. In this commonly
used invasive blood pressure monitoring setup, a catheter is inserted into an
artery (or a vein if venous pressure is monitored), and an arterial pressure
extension tube filled with saline is connected to the catheter. This setup is
often referred to as the arterial line. The other end of the extension tube is
connected to a pressure transducer. The transducer, which converts the pres-
sure signal to an electrical signal, is connected to a pressure monitor to dis-
play the blood pressure waveform and the systolic, mean, and diastolic pres-
sure values. To prevent blood clot at the tip of the catheter inside the blood
vessel (blood clot will block the pressure signal from reaching the transduc-
er), a bag of heparinized saline is connected to the extension tube. The bag
is pressurized (to about 150 mmHg) to above the blood pressure in the ves-
sel. Together with the continuous flush valve at the transducer set, this setup
produces a slow continuous flow of heparinized saline flushing the catheter
to prevent blood clot. The flow is very slow (less than 5 ml/hr) to avoid cre-
 Invlarive Blood Pressure Monitors

ating a pressure drop in the extension tube and catheter setup; otherwise it
will affect the accuracy of blood pressure measurement. The rapid flush
valve is used during initial setup to flush and fill the extension tube before it
is connected to the indwelling catheter.

Heparinized

F Rapid flush
button Cable to IBP
Continuous
flush valve\ \ /
monitor

Pressure
transducer

3-way valve

/
Flush valve
and disposable
transducer set

To indwelling m-.
catheter

Figure 18-7. Arterial Blood Pressure Monitoring Setup.

The equivalent hydraulic circuit of the arterial line setup is shown in

Figure 18-8a. In the setup, the patient port (which is the location of the
catheter) is h meter above the transducer port (the point where the liquid in
the extension tube interfaces with the pressure transducer). Therefore, the
pressure Px as seen b y the transducer is the sum of the pressure due to the
liquid column and the pressure PPat the patient port (blood pressure of the
patient), i.e.,
Px = PP + pgh, (1)
where p is the density of the liquid in the extension tube and g is the accel-
eration due to gravity. This equation shows that the pressure as seen by the
transducer is higher than the actual blood pressure by the product pgh. Figure
18-8b shows the concept of introducing an offset Po to compensate for this
overpressure phenomenon. The zeroing process performed during the initial
setup is designed to allow the machine to determine the value of this offset.
 Biomedical Device Technology: Pfinc@lesand Design

Patient port
(PP)

L:~m-y~
transducer

Po
ha:
display
i

Offset
voltage

(a) (b)
Figure 18-8. Zeroing of Blood Pressure Monitor.

From the compensation circuit and equation (I),

P = Px + Po = PP + pgh + Po.
During the initial zeroing process, the patient port is open to the atmos-
phere so that PP becomes zero. Equation (2) therefore becomes
P = pgh + Po. (3)
While the patient port is still exposed to atmospheric pressure, the oper-
ation invokes the zeroing sequence of the blood pressure monitor telling the
monitor that this reading corresponds to zero pressure (i.e., P= atmospher-
ic pressure = 0, or zero gauge pressure). Equation (3) now becomes
0 = pgh + Po Po = - pgh.
The blood pressure monitor saves this value of Po in memory and exits
the zeroing sequence. The operator then closes the patient port. The blood
pressure monitor applies this offset to the transducer reading during subse-
quent pressure measurements. As long as the vertical height difference
between the transducer port and patient port remains the same, the monitor
will always display the true blood pressure of the patient. However, if the
transducer is lowered (or the patient port is raised) after zeroing, the monitor
will over-read the pressure (for every 2.5 cm decrease in the height differ-
ence, the pressure is over-read by 2 mmHg). The zeroing process will also
compensate for any other constant offsets in the system, including offset from
the pressure transducer.
 Invasive Blood Pressure Monitors

Example
A patient is undergoing invasive blood pressure monitoring. the arterial line
was zeroed at setup. The patient's systolic pressure and diastolic pressure
were 125 and 80 mmHg, respectively. If the patient bed is raised by 4.0 inch-
es while the level of the transducer remains the same, what will the pressure
readings be?

Solution
Using the equation Po = pgh and assuming the density of the saline in the
extension tube has a density of 1,020 Kg/m3, raising the patient by 4 inches
(4.0 x 0.0254 = 0.10 m) will increase the offset pressure by

The blood pressure reading therefore becomes 132.5 and 87.5 mmHg.

In most prepackaged disposable transducers, the zero ports are attached

to the transducers. In order to properly zero the system, the setup procedure
requires that the zero port be located at the same level as the patient's heart.
Instead of opening the patient port to the atmosphere during the zeroing
process, the zero port attached to the transducer is open to the atmosphere.
In this case, after correct zeroing, the monitor will display the blood pressure
at the level of the patient's heart.

FUNCTIONAL BUILDING BLOCKS OF AN INVASIVE BLOOD

PRESSURE MONITOR

Figure 18-9 shows a typical functional block diagram of an invasive

blood pressure monitor. The following paragraphs describe the functions of
each building block.

The pressure transducer of an invasive blood pressure monitor converts

the pressure signal to an electrical signal. Ideally, the transducer should have
a linear characteristic with adequate frequency response to handle the rate of
pressure fluctuations.
Figure 18-10a shows the cross-sectional view of a four-wired resistive
strain gauge pressure transducer. The central floating block is connected to
the four pretensioned strain wires with the other ends of the wires connect-
 Biomedical Device Technology: Principles and Design

I 1

Input Blood pressure Zero offset

pressure transducer control

Transducer Amplifiers and

excitation filters

Signal isolation

Signal
processing

Waveform and
numeric display

Figure 18-9. Functional Block Diagram of an Invasive Blood Pressure Monitor.

ed to the stationary frame of the transducer. The diaphragm in contact with

the fluid chamber (or pressure dome) is mechanically connected to the cen-
tral floating block by a rigid connecting rod. The blood pressure to be mea-
sured is transmitted via saline in the extension tube to the fluid chamber,
forcing the diaphragm to move according to the pressure fluctuation. As the
diaphragm moves, the strain wires will extend or retract according to the
movement. The strain wires are connected in a bridge format with the elec-
trical equivalent diagram as shown in Figure 18-10b.
For a higher applied pressure, strain wires 1 and 2 are stretched while 3
and 4 are relaxed. As a result, the resistance of the strain wires 1 and 2
(shown in the equivalent circuit) become higher while 3 and 4 become less.
If an excitation voltage VE is applied to the bridge, the output voltage T/'o will
vary according to the applied pressure. Although VE is shown to be a DC
voltage, AC excitation may be used. The sensitivity of a pressure transducer
is often expressed in outpbt voltage per unit pressure (e;g, 10 mV/mmHg)
or in output voltage per unit excitation voltage per unit pressure (e.g., 2
mV/V/mmHg). -
In older systems or systems using reusable transducers, resistive strain
gauge or piezbelectric eliment transd;cers are used. Nowadays, disposable
transducers using semiconductor piezoresistive strain gauges are commonly
used. These transducers are mass-produced using semiconductor fabrication
 Inuasive Blood Pressure Monitors

Pressure sensing

(b) Electrical Equivalent Circuit

(a) Transducer Construction

Figure 18-10. Resistive Strain Gauge Pressure Transducer.

technology, which yields consistent performance at low cost. Disposable IBP

transducers are prepackaged with the extension tube, flush valves, and con-
nectors for single patient use.
The sensitivity of a disposable blood pressure transducer as specified in
the AAMI Standards BP22 and BP23 is 5 pV/V/mmHg f 1% when an exci-
tation voltage of 4 to 8 V, D C to 5 kHz is used. This allows interchangeabil-
ity of blood pressure transducers across different manufacturers of compati-
ble blood pressure monitors.

Example
A special-purpose reusable pressure transducer has an output sensitivity of
2.0 mV/V/mmHg. If an applied pressure of P= 200 mmHg is applied and
the excitation VEis a 5.0 V peak to peak 100 Hz sinusoidal voltage source,
what is the output voltage of the transducer?
 Biomedical Device Technoloa: Princ$les and Design

Solution
Since
Vo = sensitivity x VE X P.
The transducer output voltage Vo is calculated by:
Vo = 2.0 mV/V/mmHg X 5.0 VPp X 100 mmHg = 1,000 mVpp or 1.0 Vp-p
The output voltage in this case is also 100 Hz sinusoid.

Zero Offset
The zero offset functional block is used during the zeroing procedure to
determine and store the zero offset value. During blood pressure monitoring,
this stored value is used to compensate for the offset due to the static pres-
sure of the setup and the offset of the transducer circuit. For microprocessor-
based machines, the zero offset value is stored digitally.

Amplifiers and Filters

For a systolic pressure of 120 mmHg, a standard disposable transducer
(sensitivity = 5 pV/V/mmHg ) with an excitation of 5 V produces an output
of 3,000 pV or 3 mV; such small voltage must be amplified before it can be
used by other parts of the monitor. Instrumentation amplifiers with high
input impedance and high common mode rejection ratio are used for this
purpose.
The fundamental frequency of the blood pressure waveform is the same
as the heart rate (which is about 1 Hz). However, spectral analysis of a typi-
cal arterial blood pressure waveform (Figure 8-1 lb) shows a bandwidth from
DC to about 10 Hz (Figure 8-1 lb). In order to reduce higher frequency arti-
facts and interferences, a low pass filter with a high cutoff frequency of 20 to
50 Hz is often built into the front-end analog circuit of the monitor.

Signal Isolation
Invasive blood pressure monitoring requires external access to major
blood vessels. Both the saline solution in the arterial line and the blood in the
artery conduct electricity. This setup forms a conduction path between the
electromedical device and the patient's heart. The blood pressure monitor
may become the source or sink of the risk current flowing through the heart.
Signal isolation to break the conduction path is required to minimize the risk
of electrical shock (both macro- and microshocks) to the patient.
 Invasive Blood Pressure Monitors

Pressure (mmHg)

- I 2

(a) Time Domain Waveform

Time (s) 10
I
20

(b) Frequency Spectrum

Figure 18-1 1. Frequency Spectrum of Blood Pressure Waveform.
b
Freq (Hz)

Signal Processing
From the blood pressure waveform, the systolic, mean, and diastolic
blood pressures are determined. In addition, the patient's heart rate can be
derived as the frequency of the pressure cycle is the same as the cardiac
cycle. The systolic blood pressure is obtained by using a peak detector cir-
cuit (Figure 18-12a). In order to track the fluctuating systolic pressure, a pair
of peak detectors arranged in a sample and hold configuration are used. The
diastolic pressure can be found by first inverting the pressure waveform and
then finding the peak of this inverted waveform. Mean blood pressure is
obtained using a low pass filter circuit (Figure 18-12b).
In modem monitors, the blood pressure waveform is sampled and con-
verted to digital signal. Systolic, mean, and diastolic pressures are deter-
mined by software algorithm in the microprocessor.

Display
Liquid crystal displays (LCDs) have been replacing cathode ray tubes
(CRTs) in recent years as the display of choice for medical waveform dis-
plays. In addition to waveforms, numeric information is also displayed on
the medical monitor.
 Biomedical Device Technoloa: Pfinc$les and Design

Peak detector PSYS~OI~C

(a) Analog Peak Detector 1-

LOWpass filter P~ean

(b) Low Pass Fllter

(c) Systolic, d~astolic,and mean blood pressure detector
Figure 18-12. Blood Pressure Detection from Pressure Waveform.

COMMON PROBLEMS AND CAUSES OF ERRORS

Problems in an invasive blood pressure measurement system may pro-

duce inaccurate pressure readings or distorted blood pressure waveform.
These erroneous signals may cause improper diagnosis, leading to inappro-
priate medical intervention. Some common sources of errors are described
next.

Setup Error
The most common problem in this category relates to the zeroing
process. Incorrect zeroing procedure or change in vertical distance between
the transducer and measurement site after initial setup produces a constant
static pressure error in the measurement. It is important for the clinician to
correctly perform the zeroing procedure, understand the principles, and be
aware of the implications from setup variations.

Catheter Error

Although there is no active component in the catheter and it seems to be

a very simple part of the blood pressure monitoring system, many artifacts
and measurement errors may arise from the catheter. Some of the common
 Invasive Blood Pressure Monitors

problems are described next.

End pressure, whipping, and impact artifacts-The catheter in blood pres-
sure monitoring is a small flexible tube inserted into a blood vessel with pul-
sating blood flow (Figure 18-13). If the blood is flowing in the same direction
of the catheter, it creates a small negative pressure at the end of the catheter
tip. In contrast, blood flowing toward the catheter tip will create a net posi-
tive pressure. The flow of blood may create turbulence and set the catheter
tip into whipping motion. Movement of the catheter may cause it to collide
with the vessel wall or valves. Whipping motion and impact of the catheter
may show up as distortion in the blood pressure waveform.

Catheter

To transducer
Blood flow

Blood vessel
Figure 18-13. Catheter in Blood Vessel.

Air bubble, pinching, and leak-Another area of pressure waveform dis-

tortions is caused by the reduction of the frequency response of the catheter-
extension tube. An air bubble in the catheter or fluid-filled extension tube
reduces the cutoff frequency of the high pass filter formed by the hydraulic
circuit. Pinching the line or having a leak in the line has a similar effect. Such
problems in the system will attenuate the high-frequency component of the
blood pressure waveform. Figure 18-14 shows the effect of such problems on
the frequency response of a catheter.

Blood Clot
The purpose of the pressured inhsion bag is to prevent a blood clot
occurring at the tip of the catheter in the blood vessel. A total blood clot will
block the transmission of the pressure signal to the transducer. A blood clot
will diminish the amplitude fluctuation (difference in systolic and diastolic
pressure) and lower the high-frequency response of the setup. Periodic
 Biomedical Device Technology: Principles and Design

Normalized Air bubble

amplitude 4

r
I
100 Frequency (Hz)
Figure 18-14. Frequency Response of Catheter Setup.

inspection of the drip chamber attached to the infusion bag to ensure a con-
tinuous flow of the heparinized saline will prevent clotting.

Transducer Calibration
Due to stringent manufacturing processes, there is no need to perform
field verification of the accuracy of single-use disposable blood pressure
transducers. However, blood pressure monitors must be checked periodical-
ly to ensure that they are functioning properly with amplification and fre-
quency response according to manufacturers' specifications. In practice, a
simulator is used to provide a known input to the monitor and the output is
measured and compared with the specifications.
For reusable pressure transducers, a known pressure source is used to
determine the sensitivity of the transducer. Most pressure monitors have a
calibration factor (F)adjustment to compensate for sensitivity drift of the
transducers. A simple procedure to obtain the calibration factor of a particu-
lar transducer is:
1. Apply a known pressure (I?)to the transducer and read the pressure
display (Pd)on the monitor.
A .
2. Calculate the calibration factor by using the equation F = -
Pd
3. Input the value of Finto the calibration factor adjustment input of the
monitor.
4. The monitor is now calibrated to use with this particular transducer.
 Inuasiue Blood Pressure Monitors

Example
A 200 mmHg pressure source is used as input to determine the calibration
factor of the monitor with a reusable pressure transducer. If the pressure
reading of the monitor is 190 mmHg, what is the calibration factor?

Solution

Using F= -Pi , the calibration factor is F = -

2oo - 1.05.
Pd 190

Hardware Problems
As with all medical devices, there is always a possibility of component
failure. It is important that users be able to differentiate between normal and
abnormal performance of the monitoring system. Many monitors have built-
in simple test procedures to allow the users to verify the function and per-
formance of the system. In order to ensure that the monitor is functioning
according to standards or manufacturers' specifications, periodic perfor-
mance verification inspections by qualified professionals are required to
detect nonobvious problems such as component parameter drifts.
 Chapter 19

NONINVASIVE BLOOD
PRESSURE MONITORS

OBJECTIVES

Identify the components of a sphygmomanometer.

Describe the principles of operation and the limitations of using a man-
ual auscultatory method to measure systolic and diastolic blood pres-
sure.
Differentiate the auscultatory and oscillometric methods employed in
automatic NIBP measurement.
Describe the principles of operation and limitations of NIBP measure-
ment using the oscillometric method.
Identify the functional building blocks of a typical oscillometric NIBP
monitor.
Explain the principles of using Doppler ultrasound and tonometry in
NIBP monitoring.

CHAPTER CONTENTS

1. Introduction
2. Auscultatory Method
3. Oscillometric Method
4. Other Methods of NIBP Measurement
 Noninvasive Blood Pressure Monitors

INTRODUCTION

Blood pressure, an important physiological parameter, is measured rou-

tinely throughout the course of nearly all medical procedures and diagnosis.
While one may measure blood pressure accurately using the invasive tech-
nique described in the previous chapter, being able to measure blood pres-
sure noninvasively is a tremendous achievement made possible by the com-
bination of creativity and technology.
In 1876, E. J. Marey, a French physiologist, performed the first experi-
ment using counterpressure to measure blood pressure. Marey was investi-
gating the interaction between the arterial pulsatile pressure with an applied
external pressure. He had his assistant place his hand into a jar filled with
water; the jar was sealed at the wrist. The pressure in the jar was increased
incrementally and the pressure oscillation in the jar was measured at each
pressure increment. Marey noticed that when the jar pressure was slightly
above the systolic pressure, all the blood was driven out of the hand and no
pressure oscillation was detected. He further noticed that the amplitude of
oscillation began to rise as the jar pressure dropped below the systolic pres-
sure, reached at maximum, and decreased as the pressure was further
reduced. This method has since evolved into the oscillometric method of
noninvasive blood pressure (NIBP) measurement-the most popular method
to noninvasively measure blood pressure. The auscultatory method was
introduced in 1905 by J. S. Korotkoff.
Although results from NIBP measurement may not be as accurate as
invasive methods, NIBP measurement is easy, nonhazardous, and inexpen-
sive. It provides safe and reliable trending of a patient's blood pressure in
clinical settings. Today, NIBP measurement is performed in almost every
medical examination. This chapter describes the principles and instrumenta-
tions of a number of common indirect methods in blood pressure measure-
ment.

AUSCULTATORY METHOD

Indirect methods measure blood pressure without directly accessing the

bloodstream. The most commonly used instrument is based on the ausculta-
' tory technique. The device used in this technique is called a sphygmo-
manometer, which is present in every hospital bedside, clinic, and physi-
cian's office.
-
A sphygmomanometer (Figure 19-1) consists of:
I 1. An inflatable rubber bladder enclosed in a fabric cover called the
P
 Biomedical Device Technology: Princ$les and Design

cuff;
2. A rubber hand pump with valve assembly so that the pressure in the
setup can be raised and released at a slow controlled rate; and
3. A pressure measurement device. Mercury manometers were com-
monly used as the pressure measurement device. However, since
mercury is a hazardous material, rotary mechanical air pressure
gauges have replaced mercury manometers to measure the pressure
in the cuff.
In addition to the sphygmomanometer, a stethoscope is required to lis-
ten to the sounds in the artery during the measurement.
During blood pressure measurement, the pressure cuff is wrapped
around the upper arm of the subject and a stethoscope is placed on the inner
elbow for the operator to listen to the sound produced by the blood flow in
the brachial artery. While watching the pressure gauge, the operator manu-
ally squeezes the hand pump to raise the cuff pressure until it is above the
systolic blood pressure (e-g., 150 mmHg). At this pressure, the brachial artery
is occluded. Since blood is not able to flow to the lower arm, no sound will
be heard from the stethoscope. The cuff pressure is then slowly reduced, say,
at a rate of approximately 3 mmHg per second, by opening the pressure
release valve. As the cuff pressure falls below the systolic pressure, the clini-
cian will start to hear some clashing, snapping sounds from the stethoscope.
This sound is caused by the jets of blood pushing through the occlusion. As
the cuff pressure continues to decrease, the sound intensity will first increase;
it will then turn into a murmur-like noise and become a loud thumping
sound. The intensity and pitch of the sounds will change abruptly into a muf-
fled tone when the cuff pressure is getting close to the diastolic pressure and
will disappear completely when the pressure is below the diastolic pressure.

Figure 19-1. NIBP Measurement Setup Using Manual Auscultatory Method.

 Noninvasive Blood Pressure Monitors

These sounds are called Korotkoff sounds.

The cuff pressure at which the first Korotkoff sound appears corresponds
to the systolic pressure, while the disappearance of the sound corresponds to
the diastolic pressure. Figure 19-2 shows the relationships between the arte-
rial pressure and the cuff pressure during the course of measurement. This
method is suitable for most patients, including hypotensive and hypertensive
patients. Other than applying the cuff over the upper arm, the cuff can be
placed over the thigh or the calf of the patient. In each of these applications,
the Korotkoff sounds should to be detected downstream of the occlusions.

Pressure 4

Cuff pressure

Time
Figure 19-2. Relationships Between Arterial and Cuff F'ressure.

The accuracy of this method has several limitations:

The Korotkoff sounds are normally in the range of less than 200 Hz,
where human hearing is normally less acute. Determination of the
Korotkoff sounds is affected by the hearing acuity of the operator,
especially when it is used on hypotensive patients or infants, where the
sound levels are low.
Inappropriate cuff size or incorrect placement can produce falsely
high (undersized or loosely applied cuff) or falsely low (oversized cuff)
readings. The American Heart Association (AHA) recommends that
the length of the bladder under the cuff be 80010 of the circumference
of the patient's limb and the width of the bladder be 40% of the cir-
cumference.
A too fast rate of cuff deflation will produce an erroneous reading.
 Biomedical Device Technology: Principles and Design

Figure 19-3 shows an underestimation of the systolic pressure due to

a too fast deflation rate.
It is known that the Korotkoff sounds disappear early in some patients
and then reappear as the cuff pressure is lowered toward the diastolic
pressure. This phenomenon is referred to as the auscultatory gap.
The auscultatory method can determine the systolic and diastolic pres-
sures but not the mean blood pressure.
Although the sphygmomanometer is a relatively simple device, regular
maintenance is still required, which includes pressure gauge calibration;
cleaning; and checking for leaks on tubing, cuff bladder, valves, and the hand
pump.

Arterial pressure

Cuff pressure

Time
Figure 19-3. Error in NIBP Measurement on Fast Deflation.

An automatic NIBP monitor uses the same principle as the manual aus-
cultatory method. Automation overcomes the hearing acuity limitation by
employing a microphone inside the cuff to pick up the Korotkoff sounds
instead of relying on human hearing. It also replaces the manual pump with
an automatic pump and uses an electronic pressure transducer instead of a
mechanical pressure gauge. After the cuff is applied, the NIBP monitor auto-
matically inflates the cuff to occlude the blood vessel. The bladder pressure
is slowly released while the microphone listens for the Korotkoff sounds.
These processes are automatically coordinated by the monitor. The systolic
and diastolic pressures are determined by tracking the bladder pressure and
correlating it to the different phases of the Korotkoff sounds picked up by the
microphone.
 Noninvasive Blood Pressure Monitors

OSCILLOMETRIC METHOD

NIBP monitors using the oscillometric method are similar to the auscul-
tatory method except the oscillometric method detects the small fluctuations
of pressure inside the cuff rather than listening to the Korotkoff sounds in the
auscultatory method. When the cuff pressure falls below the systolic pres-
sure, blood breaks through the occlusion, causing the blood vessel under the
cuff to vibrate. This vibration of the vessel's wall causes fluctuation (or oscil-
lation) of the cuff pressure. The onset of the vibration correlates well with the
systolic pressure, while the maximum amplitude of oscillation corresponds
to the mean arterial blood pressure. When the cuff pressure is at the mean
arterial pressure, the net average pressure on the arterial wall is zero (both
sides of the wall are of the same pressure), which allows the arterial wall to
freely move in either direction. Under this condition, the amplitude of vibra-
tion of the arterial wall caused by blood pressure fluctuation in the artery is
the highest. The diastolic pressure event on the oscillometric curve is some-
what less defined. One commonly adopted approach to determine the dias-
tolic pressure is to take the point where the amplitude of the oscillation has
the highest rate of change; another approach estimates the diastolic pressure
by locating the point where the cuff pressure corresponds to a fixed per-
centage of the maximum oscillation amplitude.
Figure 19-4a shows the relationships between the arterial blood pressure
and the cuff pressure. The maximum amplitude of pressure oscillation is usu-
ally less than a few percent of the cuff pressure. To extract only the oscilla-
tory component from the pressure signal obtained by the pressure sensor, the
low-frequency component of the signal (corresponding to the slowly deflat-
ing cuff pressure) is removed by a high pass filter. The remaining oscillatory
component of the signal (shown amplified in Figure 19-4b) is then used to
determine the mean, systolic, and diastolic blood pressures. The cuff pres-
sure corresponding to the maximum oscillation amplitude is taken as the
mean arterial pressure. Different manufacturers of NIBP monitors may use
different algorithms to determine the systolic and diastolic pressures from
this oscillometric signal.
Compared to the auscultatory method, NIBP measurements using the
oscillometric method are not affected by audible noise and therefore can
work in a noisy environment. On the other hand, as this method relies on
detecting the amplitude of pressure fluctuation, any movement or vibration
can lead to incorrect readings. Furthermore, in oscillometric NIBP monitors,
the diastolic pressure is only an estimated quantity. In addition, the small
pressure change at the onset of oscillation (which corresponds to the systolic
pressure) is difficult to detect. Of the two automatic noninvasive methods,
the oscillometric method is more commonly used than the auscultatory
 Biomedical Device Technology: Principles and Design

Pressure A

Ld Cut7 pressure

Time

High pass filtered and

amplified cuff pressure
Maximum pressure
fluctuation
Figure 19-4. Relationships Between Arterial
and Cuff Pressure in Oscillometric Method.

method in automatic blood pressure monitors.

Figure 19-5 shows the functional building blocks of a NIBP monitor
using the oscillometric method. The following descriptions explain the func-
tions of the building blocks.
Pump and Solenoid klve. The motorized air pump inflates the cuff pres-
sure to a predetermined pressure so that the artery is occluded under the
cuff. The solenoid valve connects the air circuits between the pump and the
cuff during inflation and connects the cuff to atmosphere during deflation.
The rate of deflation can be controlled by pulsing the solenoid at a certain
duty cycle.
Pressure Sensor. Through the internal tubing connections, the cuff pressure
is constantly monitored by the pressure sensor in the NIBP monitor.
Amp1iJ;ers and Oscillometric Filter. The signal picked up by the pressure
transducer is amplified (by about 100 times). This signal consists of two sets
of information: the slowly decreasing cuff pressure and the oscillatory signal.
The slow varying signal is separated from the oscillatory signal by a high pass
filter (with cutoff frequency of about 1 Hz). Both signals are fed to the ana-
log to digital converter.
 Noninvasive Blood Pressure Monitors

filter and

Analog to digital
Amplifier
converter

$.
Pump/solenord Memoly
control

Patrent safety
circuitry
D~splayand
V pnnter
Motor speed
controller
Network
4b ~nterface
Overpressure
svvrtch

-
User Input
Pressure
Pump
sensor

To cuff f= - Solenoid
valve
Watchdog

electrical circuit air circuit

Figure 19-5. Functional Block Diagram of Oscillometric NIBP Monitor.

Central Processing Unit (CPU) and Analog to Digital Converter (ADC). The
cuff pressure and oscillometric signal are digitized by the ADC and sent to
the CPU to determine the mean, systolic, and diastolic pressures of the mea-
surement. The heart rate can also be determined from the signals.
Display, Printer, Memory, and Network Interface. The measured systolic,
diastolic, and mean blood pressures are shown on a display (e.g., LCD). A
hard copy may be printed for charting. These data may also be time-stamped
and saved in the memory of the monitor for trending or communicated via
network connections to other devices.
Watchdog Timer and Overpressure Switch. An independent overpressure
safety switch activates the solenoid valve to release the cuff pressure down to
atmospheric pressure should excessive pressure develop in the cuff. The
solenoid will also open to the atmosphere if the cuff pressure remains high
 Biomedical Device Technology: Principles and Design

for a preset duration of time. Both features are in place to prevent compres-
sion damage of the tissues under the cuff.

OTHER METHODS OF NIBP MEASUREMENT

There are many other methods to measure or estimate blood pressure

noninvasively. Compared to the auscultatory or oscillometric methods, these
methods are either not as accurate or more complicated or not as easy to use
in clinical settings. Two of the better methods are described next.

Doppler Ultrasound Blood Pressure Monitor

This class of device makes use of the Doppler effect to detect blood flow
patterns in the artery of interest. A sound transmitter and a receiver are
placed inside the pressure cuff. The monitor detects the Doppler shift when
the incident sound wave is reflected from the blood flow in the subject.
When the artery is occluded by the cuff, the Doppler shift is zero. When the
cuff pressure is slowly reduced, the arterial pressure is able to overcome the
cuff pressure occlusion, causing the occlusion to snap open. This jet of blood
flowing through the cuff occlusion produces a Doppler shift. There are actu-
ally two Doppler events during each cardiac cycle-the opening and closing
of the blood vessel under the cuff. When the arterial pressure exceeds the
cuff pressure, the blood flowing through the opening of the occlusion pro-
duces a high-frequency Doppler shift (e.g., 200 to 500 Hz). When the arteri-
al pressure recedes toward the diastolic pressure, the blood vessel will be
reoccluded. This event produces a lower frequency Doppler shift (e.g., 15 to
100 Hz).
When the cuff pressure is allowed to be bled down at constant speed
from the systolic pressure, the high- and low-frequency events appear and
are next to each other. As the cuff pressure continues to drop, the two events
become farther and farther apart. When the cuff pressure reaches the dias-
tolic pressure, the low-frequency event will coincide with the high-frequency
event of the next cardiac cycle (Figure 19-6). The frequency shift may be
coupled to a loudspeaker to allow the operator to determine the systolic and
diastolic pressures. Figure 19-6 shows the Doppler events due to the inter-
action of the cuff pressure and blood pressure.
 Noninvmive Blood Pressure Monitors

Cuff pressure Arterial pressure

Time
High-frequency boppier Low-frequency Doppler
(opening event) (closing event)

Figure 19-6. Doppler Events Due to Interaction of Cuff Pressure and Blood Pressure.

Arterial Tonometry
None of the NIBP methods discussed are able to measure the blood pres-
sure waveform. Arterial tonometry is a continuous pressure measurement
technique that can noninvasively measure pressure in superficial arteries
with sufficient bony support, such as the radial artery. A tonometer is a con-
tact pressure sensor that is applied over a blood vessel. It is based on the
principle that if the sensor is depressed onto the vessel wall of an artery such
that the vessel wall is parallel to the face of the sensor, the arterial pressure
is the only pressure perpendicular to the surface and is measured by the sen-
sor (Figure 19-7). Theoretically, accurate real-time blood pressure waveform

Sensor

Figure 19-7. Arterial Tonometry.

 Biomedical Device Technology: Pfinc$les and Design

can be recorded using this noninvasive technique. However, experiments

showed that although this method produces good-quality pulse waveform, it
tends to underestimate the systolic and diastolic pressures.
To obtain good results, tonometry requires that the contact surface be stiff
and the sensor be small relative to the diameter of the blood vessel. In addi-
tion, proper sensor application is critical because if the vessel is not flattened
sufficiently (e.g., due to inadequate depression), the tonometer will measure
forces due to arterial wall tension and bending of the vessel. However, too
much depression force may occlude the blood vessel.
 Chapter 20

CARDLAC OUTPUT MONITORS

Define the terms cardiac output, stroke volume, and cardiac index.
State the Fick principle and the indicator dilution method.
Describe how to measure cardiac output using oxygen and heat as the
"tracer."
Explain the principle of the thermal dilution method in cardiac output
measurement.
Review the setup and the procedures to measure cardiac output using
the thermal dilution method.
Sketch the block diagram of a cardiac output monitor using the thermal
dilution method.
Identify potential sources of error in cardiac output measurement and
methods to minimize errors.

CHAPTER CONTENTS

1. Introduction
2. Definitions
3. Direct Fick Method
4. Indicator Dilution Method
5. Thermal Dilution Method
6. Problems and Errors
 Biomedical Device Technology: Principles and Design

INTRODUCTION

Cardiac output is a measurement of the performance of the heart. It is

also used to calculate many hemodynamic functions. The heart serves as a
pump to circulate blood around the cardiovascular system. In fluid mechan-
ics, the power produced by a pump is determined by its output pressure and
volume flow rate. In the last two chapters, we have studied devices to mea-
sure blood pressure. In this chapter, we are going to study cardiac output
monitor-a medical device to measure blood flow of the heart.
Although there are many direct and indirect methods to measure cardiac
output, since the introduction of the Swan-Ganz catheter in the 1970s, the
thermal dilution method has become a standard procedure to measure car-
diac output in intensive care units, surgical suites, and cardiac care units.
The thermal dilution method in cardiac output measurement is an appli-
cation of the indicator dilution method based on the Fick principle. The Fick
principle was proposed by Adolf Fick and states that the rate Qof a substance
delivered to an area with a moving fluid stream is equal to the product of the
flow rate Fof the fluid and the difference in concentration Cof the substance
at sites proximal and distal to the area. In equation format: Q= F (Cd - Cp)
or:

DEFINITIONS

For every contraction, the heart pushes a certain volume of blood into
the common aorta. This volume of blood pumped by the ventricles during
one ejection is defined as the stroke volume or SV. Therefore, the volume of
blood pumped out from the heart per unit time is equal to the SV multiplied
by the heart rate or HR. This product, which is the volume of blood pumped
out by the heart per unit time, is defined as the cardiac output. Cardiac out-
put or C O is commonly expressed in liters per minute (L/min). Therefore,
C O = SV X H R where SV is in liters, and HR is in beats per minute.
For a normal adult, the resting C O is about 3 to 5 L/min. However, dur-
ing intense exercise, since both the H R and SV become higher, the C O of
the same individual may have increased to several times of that at rest (e.g.,
to 25 L/min). As with all physiological signals, the resting C O varies from
person to person and is often dependent on body size. To facilitate compar-
ison, C O is often normalized by dividing it by the weight or by the body sur-
face area of the patient. The latter is called the cardiac index, which has a
 Cardiac Output Monitors

unit of Ymin/m2. A typical resting cardiac index is 3.0 L/min/m2. One may
wonder how body surface area is determined. In fact, lookup tables of body
surface area based on the weights and heights of typical individuals are avail-
able. Alternatively, an empirical formula can be used to obtain the body sur-
face area:
(2)
where A = total body surface area in m2,
W = body weight in kg, and
H = height in cm.
For example, the body surface area A of a 70 kg, 1.7m tall patient is
A = 70 I1 425 X 170 O n 5 X 0.007184 = 1.73 m2.

DIRECI' FICK METHOD

The direct Fick method is considered to be the "gold standard" in car-

diac output measurement. It uses oxygen as the indicator and assumes that
the left ventricular blood flow is equal to the blood flow through the lungs.
This method involves measurement of the rate of oxygen uptake of the lungs
and the oxygen content of the arterial blood and venous blood. Deoxy-
genated blood from the right ventricle, which has the lowest oxygen con-
centration in the cardiovascular system, enters the lungs and picks up oxy-
gen to become oxygenated blood. The oxygenated blood then flows via the
left atrium, left ventricle, and common aorta into the arteries. According to
the Fick principle, the volume blood flow Fthrough the lung (i.e., the cardiac
output) can be obtained if the rate of oxygen uptake Qand the difference in
oxygen concentration Ca of the arterial blood and oxygen concentration Cu
of the venous blood are known. Using these quantities, equation (1) then
becomes:

F=Q
(Ca - Cv) '
In practice, blood samples are drawn during measurement of oxygen
consumption. The venous blood is drawn from the pulmonary artery and the
arterial sample is taken from one of the main arteries. Oxygen content of the
venous and arterial blood is determined by laboratory analysis of these
blood samples. The oxygen consumption Qis calculated from the rate of gas
inhalation and the difference of the oxygen concentrations in the atmos-
pheric air and the expired air from the patient. The rate of gas inhalation is
measured using a spirometer and the expired gas oxygen concentration is
 Biomedical Device Technology: Principles and Design

measured using an oxygen analyzer. The blood flow rate F, or cardiac out-
put, is then calculated. In this method, the subject must be in a steady state
throughout the period of measurement (about 3 minutes) to avoid transient
changes in blood flow or in the rate of ventilation.

Example
In a cardiac output measurement using the direct Fick method, the rate of
oxygen consumption was found to be 300 mVmin. Blood sample analysis
shows the arterial and mixed venous oxygen contents are 200 ml/l and 140
ml/l, respectively. Calculate the cardiac output.

Solution
Using equation (3),

INDICATOR DILUTION METHOD

The indicator dilution method is a variation of the Fick principle. The

indicator dilution method to measure fluid flow rate is based on the upstream
injection of a tracer (or detectable indicator) into a mixing chamber and mea-
suring the concentration-time curve (or dilution curve) of the tracer down-
stream of the chamber (Figure 20-la). To obtain accurate results, the tracer
is required to thoroughly mix with the fluid in the mixing chamber.
Theoretically, for the same flow and same tracer injection volume, the area
under the dilution curve will be the same even though the shapes of the
curves are different. Figure 20-lb shows the ideal indicator dilution curve
obtained by immediate mixing of the tracer with the fluid after injection and
having the same fluid velocity over the entire cross section of the tube. In the
ideal case (Figure 20-lb), the fluid flow rate Fis proportional to the amount
of tracer m injected and inversely proportional to the concentration C of the
tracer and the duration Tof the concentration curve, or simply:

Note that the product C and Tis the area under the dilution curve.
 Cardiac Output Monitors

Tracer inject~on 4 Mixing chamber Tracer concentration

at prox~malend detect~onat distal end

Fluid flow a) Setup

b) ldeal~zedd~lutioncurve

Time

c) Typical d~lutioncurve

Time
t-----T-
Figure 20-1. Indication Dilution Method.

Figure 20-lc shows a typical dilution curve in a realistic situation.

Although it shows a rapid rise and an exponential fall in concentration, the
area under the curve is still roughly the same as that of the idealistic curve
(Figure 20-lb) as long as the fluid flow rate and the amount of tracer inject-
ed are the same.

where A = area under the dilution curve.

Two types of indicators may be used in indicator dilution methods-dif-
fusible and nondiffusible indicators. A nondiffusible indicator will remain in
the system for a much longer period of time than a diffusible indicator. For
example, saline, which can be measured by a conductivity cell, is a diffusible
indicator in cardiac output measurement. It is estimated that over 15% of the
salt will be removed from the blood in its first pass through the lung.
Indocyanine green is a nondiffusible indicator that can be detected using
optical sensors. Experiments showed that only about 50% of it will be lost in
the first 10 minutes as it circulates around the cardiovascular system.
Measurements using a diffusible indicator tend to overestimate the cardiac
output, while recirculation of a nondiffusible indicator may result in lower
cardiac output measurements. Recirculation is the effect of increased indica-
 Biomedical Device Technology: Princ$les and Design

tor concentration when the previous bolus of indicator returns to the mea-
surement site during subsequent measurements. Figure 20-2 shows the dilu-
tion curve affected by recirculation. The dotted line shows the normal trace
of the curve if no recirculation occurs.

Dilution cutve

Recirculation

Figure 20-2. Effect of Recirculation in Indicator Dilution.

Example
In an indicator dilution method to measure cardiac output, 10 mg of indica-
tor is injected and the average concentration of the dilution curve is found to
be 2.5 mg/liter. If the indicator takes 60 seconds to pass through the detec-
tor, what is the cardiac output?

Solution
m
Using the equation F = -
CT '
m 10 mg - 4 liter - 4 l/min.
the cardiac output is -=
CT 2.5 mg/liter X 60 s 60 sec

THERMAL DILUTION METHOD

The thermal dilution method of cardiac output measurement is based on

the indication dilution method where heat is used as the indicator. In this
method, a known volume of cold solution (5%dextrose or saline) is injected
into the right atrium. This bolus of cold solution causes a decrease in the
blood temperature when it mixes with the blood in the right ventricle. The
change of blood temperature in the pulmonary artery (downstream of the
 Cardiac Output Monitors

mixing chamber) is measured to obtain the thermal dilution curve as shown

in Figure 20-3. It shows that the temperature of blood in the pulmonary
artery drops when the indicator-blood mixture passes through the tempera-
ture sensor and gradually rises back to the normal body temperature. Note
that the curve is inverted for easier reading since the cold solution causes a
negative change in blood temperature.

sec

Figure 20-3. Thermal Dilution Curve.

From the thermal dilution curve, the heat loss H f r o m the blood over
the time interval dt is:

where CB= specific heat capacity of blood,

F= blood flow rate (cardiac output),
pe = density of blood, and
Tc = temperature change (from body temperature) of the blood at time t.
The total heat loss of the blood to the injectate H i s equal to the integral
of equation (4):

where A = L!I'cdt is the area under the thermal dilution curve.

Since the total heat loss of the blood is equal to the heat gain of the injec-
tate (to raise the injectate temperature to body temperature), the heat gain of
the injectate HI can be determined from the preinjection condition of the
 Biomedical Device Technology: Principles and Design

injectate if we know the volume Vi, density pr, specific heat capacity CI and
the initial temperature TI of the injectate.
- TI).
HI = VICIPI(TB (6)
Since H= HI, equations (4) and (5) give
C B F ~ B=AVICI~I(TB
- TI)

where K = -
C1pl is a constant for a particular indicator.
CBPB
As heat (cold saline or dextrose) is a diffusible indicator, a correction fac-
tor K7 (< 1) is multiplied to equation (7) to compensate for the warming
effect of the indicator during measurement.

Example
In a cardiac output measurement using the thermal dilution method, 5 ml of
iced 5% dextrose is injected into the right atrium to obtain the thermal dilu-
tion curve. If the area under the curve is found to be 1.80°Cs and a correc-
tion factor Kr of 0.825 is used, find the cardiac output given that
c1p1
-= 1.08 for 5% dextrose and typical blood composition.
CB~B

Solution
Using 37OC as the body temperature and O°C as the initial injectate temper-
ature, from equation (9),the cardiac output is calculated by:
CO = [5 ml X 0.825 X 1.08 X (37OC - 0°C)]/1.800Cs= 91.6 ml/s = 5.51/min

In practice, a special catheter called the Swan-Ganz catheter is used to

inject a known volume (e.g., 5 ml) of cold dextrose into the right atrium of
the heart; the temperature of the blood in the pulmonary artery is measured
by a temperature sensor embedded near the distal end of the catheter to
obtain the thermal dilution curve. Mixing of indicator and blood occurs in
 Cardiac Output Monitors

the right ventricle. A Swan-Ganz catheter is a special multilumen catheter

with a distal opening, a proximal opening, a thermistor temperature sensor,
and an inflatable balloon. A typical catheter is about 110 cm long and 2 to 3
mm in diameter.
Figure 20-4 shows the construction of a four-lumen Swan-Ganz catheter,
and Figure 20-5 shows the position of the catheter in the cardiovascular sys-
tem during cardiac output measurement. The catheter is positioned into the
heart such that the injectate orifice is in the right atrium and the thermistors
in the pulmonary artery. During measurement, the injectate (cold saline or
dextrose) delivered into the right atrium is mixed with blood as it travels
through the right ventricle into the pulmonary artery. The temperature of
blood flowing through the pulmonary artery is continuously measured by a
thermistor located a few centimeters from the distal end of the catheter.
The procedures to measure cardiac output using a Swan-Ganz catheter
are:
1. Create an intravascular access to a vein (subclavian,internal jugular,
or basilic) using, for example, the Seldinger technique.
2. Insert the distal end of the catheter into the vein and slightly inflate
the balloon.

lnjectate syringe
Prox~malport
Distal lumen

Proxlmal lumen

syrlnge
Distal lumen
Proxlrnal lumen

Cross seQon of catheter

showing mulbple lumens

herm mist or
lumen

Figure 20-4. Construction of a Swan-Ganz Catheter.

 Biomedical Device Technology: Pfinc$les and Design

Pulmonary artery
Thermistor

Swan-Ganz catheter %
,/ !

l nje

Right ventricle F

Figure 20-5. Positioning of Swan-Ganz Catheter in Heart Chambers.

3. Continue to insert the catheter into the vein. The balloon will be
dragged by the blood flow to go through the right atrium, tricuspid
valve, right ventricle, pulmonary valve, and into the pulmonary
artery.
4. The position of the catheter can be estimated by the distant markings
on the catheter and verified by:
(a) using X-ray fluoroscopy (the tip of the catheter is radiopaque), or
(b) monitoring the characteristic changes in blood pressure wave-
form at the distal lumen as it travels from the vein into the heart
chambers and then into the pulmonary artery during the catheter
insertion.
5. Deflate the balloon.
6. Connect the catheter to the cardiac output monitor and initialize the
monitor.
7. Enter patient data.
8. Prepare injectate (saline or DW5-5% dextrose).
9. Measure injectate temperature (usually 0 to 5OC).
10. Inject a fixed volume of injectate (e.g., 5 ml) at a uniform rate (over
a period of 2 to 4 sec) into the injectate port.
11. The C O monitor will display the temperature change versus time
curve (thermal dilution curve) and calculate the C O using equation
(6).
 Cardiac Output Monitors

12. Several measurements (and averaging the results) may be required to

obtain a reliable reading.
A Swan-Ganz catheter setup can also be used to monitor the pulmonary
arterial wedge pressure (PAWP). The PAWP, also termed the pulmonary
capillary (wedge)pressure (PCW), reflects the mean level of pressure in the
left atrium. While the distal end of the catheter is in the pulmonary artery,
the balloon is fully inflated to occlude the pulmonary artery. The PAWP is
measured from the distal lumen by connecting a pressure transducer to the
distal port of the catheter.
Note that the setup procedure to measure blood pressure using the Swan-
Ganz catheter is the same as setting up an invasive pressure line as discussed
in Chapter 18. A C O monitor usually consists of one or more blood pressure
modules, one or more temperature modules, a computer, and interface com-
ponents. The functional block diagram of a cardiac output monitor is shown
in Figure 20-6. During a C O measurement, a blood pressure transducer BP
Txi is connected to the distal port to display the blood pressure waveform at
the distal end of the catheter. This pressure waveform is a mean to allow the
user to determine the position of the catheter during catheter insertion.
These blood pressure modules may also be used to monitor intracardiac
blood pressure. The thermistor Temp Tx 1 located near the distal end of the
catheter is use to acquire the thermal dilution curve. A second thermistor
Temp Tx 2 may be used to measure the injectate temperature. These pres-
sure and temperature signals are digitized, multiplexed, and sent to the cen-
tral processor to compute the cardiac output, cardiac index, blood pressures,
et cetera. The thermal dilution curve and the blood pressure waveform can
be viewed from the integrated display of the monitor.
Instead of using cold saline or dextrose, a nondiffusible dye (e.g., indo-
cyanine green) may be used as the indicator. The dye dilution curve is
obtained by continuously withdrawing blood from the pulmonary artery and
dye concentrate measured by a calibrated optical densitometer.

PROBLEMS AND ERRORS

Any inaccuracy in equation (9) will result in an incorrect cardiac output

measurement. Studies showed that C O measurement using thermodilution
under ideal conditions is generally within 5010 of the results obtained from
other methods. To avoid single measurement error, the usual practice is to
average the results of three good measurements. Potential error sources of
this technique are explained next.
 Biomedical Device Technology: Princ$les and Design

Waveform OIP

Numerical O/P

i----l ' anr

n t CPI
Alarm OIP

Temp Tx

I
Temp Tx 2 I
I
I

, _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.-.-. -. .- .-.-.-.-. -. . . . . . . . . . . . I

Other inputs:
patient data,
correction
factors, etc.

Figure 20-6. Functional Block Diagram of a Cardiac Output Monitor.

Catheter Dead Space and Injectate Warming

The CO equation assumes that all injectate enters the bloodstream and
heat transfer occurs only between the injectate and blood in the right atrium.
In practice, some injectate will remain in the catheter and heat is exchanged
between the injectate and the wall of the catheter during injection. To mini-
mize such errors, the first reading should be discarded because the first injec-
tion contains warm fluid in the catheter dead space. A correction factor ( K I
< 1) is multiplied to equation (8) to compensate for these errors.

Timing of Injection
Although respiratory action of the patient affects the blood pressure and
flow, it is impractical to synchronize the injection with the respiratory cycle.
Using an average of three or more sequential measurements can minimize
the variations.
 Cardiac Output Monitors

Rate of Injection
Erratic and long injection duration introduces errors in the thermal dilu-
tion curve and increases the injectate warming effect. Injection duration
should be between 2 and 4 seconds and at a steady speed.

Injectate Volume
The injectate volume should be measured accurately as this will affect
the calculation. A small syringe (e.g., 10 ml) is usually included in the pack-
age of the catheter. CO measurements using a small volume of injectate are
more likely to be affected by injectate warming.

Injectate Temperature
Theoretically, a higher volume injectate at a lower temperature should
increase the signal to noise ratio in the measurements. Studies confirmed that
cardiac output measurements using injectate at room temperature produced
higher variability than using injectate at O°C (iced injectate). It is recom-
mended that if the volume of injectate is less than 10 ml, the solution should
be iced. As well, clinicians should average more measurements if room tem-
perature injectates are used. Care should be taken to avoid warming the
injectate during handling.

Thermistor Position
If the thermistor in the catheter is in contact with the wall of the pul-
monary artery, the temperature reading will be higher, resulting in a lower
value of A in equation (8).There will be errors in the temperature measure-
ment if the thermistor is positioned inside the ventricle instead of inside the
pulmonary artery. Inside the ventricle, the injectate may not have been thor-
oughly mixed with the blood.

Frequency of Injection and Recirculation

If consecutive measurements are made within a short period of time, the
temperature of blood in the pulmonary artery may not have enough time to
return to normal body temperature before the next bolus of injectate is intro-
duced into the bloodstream. If too many measurements are done in a short
period of time, it may lower the overall blood temperature of the patient.
 Biomedical Device Technology:Princ$les and Desip

This latter problem is more significant in the dye dilution method as it takes
longer for the kidneys to remove the dye from the patient's bloodstream.

Intravenous Administration

Any intravenous fluid infusion will introduce below body temperature

fluid into the bloodstream. This effect will create errors in the thermal dilu-
tion curve.
 Chapter 21

CARDIAC PACEMAKERS

Define arrhythmia and list indications for artificial pacemaker implan-

tation.
Differentiate between endocardial and myocardial, bipolar and unipo-
lar pacing leads.
Define asynchronous, demand, and rate-modulated pacing.
Describe the procedures of pacemaker implantation.
Interpret the NBG (NASPE/BPEG) generic pacemaker codes
State applications of external pacing.
Differentiate between invasive and transcutaneous external pacing.
Describe the characteristics and design of implantable pacemakers.
Analyze the block diagram of a demand pacemaker and describe its
principles of operation.
List problems in artificial pacing.

CHAPTER CONTENTS

1. Introduction
2. Indication of Use
3. Types of Cardiac Pacemakers
4. Pacemaker Lead System
5. Implantation of Pacemaker
6. Pacing Mode Selection
7. Performance Characteristics
8. Functional Building Blocks of an Implantable Pacemaker
 Biomedical Device Technology: Principles and Design

9. Temporary Pacing
10. Potential Problems with Pacemakers

INTRODUCTION

In 1952, Paul M. Zoll, working with engineers of the Electrodyne

Company, developed a device that could stimulate the heart to contract
through large electrodes placed on the chest wall. In 1957, C. Walton Lillehei
and his colleagues, using an external pulse generator, successfully paced the
heart by directly placing electrodes in the heart muscle. In 1958, Wilson
Greatbatch developed a fully implantable pacemaker. In 1986, the U.S. Food
and Drug Administration gave market approval to a rate-responsive pace-
maker made by Medtronic.
A pacemaker is a therapeutic device designed to rectify some heart prob-
lems arising from irregular heart rhythms. It performs its function by apply-
ing controlled electrical stimulations to the heart. A healthy heart is stimu-
lated to contract by electrical impulses initiated from the sinoatrial (SA) node
located in the atrium near the superior vena cava. The SA node is also called
the natural pacemaker of the heart. An electrical impulse generated from the
SA node is conducted through the atria, causing them to contract. The
impulse eventually arrives at and depolarizes the atrioventricular (AV) node
located in the septa1 wall of the right atrium. Through the Bundle of His and
the Purkinje fibers, the action potential is distributed to the myocardium,
causing the ventricles to contract.

INDICATION OF USE

Normal sinus rhythm is a result of the continuous periodic and coordi-

nated stimulation of the atria and ventricles of the heart. Failure of any part
of this pathway will compromise the cardiac output. An arrhythmia is any
disturbance in the rhythm of the heart with respect to rate, regularity, or
propagation sequence of the depolarization wave. Depending on its nature,
an arrhythmia may be mild or life-threatening. There are two kinds of
arrhythmias:
1. Arrhythmias caused by disturbances of conduction, and
2. Arrhythmias caused by disturbances of the origin of the stimulation
Disturbances of conduction include:
Slowing of the spread of conduction of the electrical stimulation in one
 Cardiac Pacemakers

part of the conduction system, so that part of the heart is activated sig-
nificantly later than the rest, resulting in distortion of the ventricular
contraction pattern
Conduction through anomalous paths, which causes different parts of
the heart muscle to contract in an uncoordinated manner
Partial or complete blocks of the stimulation signal from the SA node
to the ventricles. Heart blocks originate from some malfunctions of the
heart's built-in electrical conduction system. The results of heart
blocks include low heart rate, heart muscle not getting enough oxygen,
and cardiac muscle becoming irritable and susceptible to irregular
rhythm. A patient with heart blocks has inadequate body oxygen and
low exercise tolerance, and in extreme cases, experiences loss of con-
sciousness and convulsion due to lack of oxygen to the brain. There
are three degrees of heart block:
1st degree-long delay in signal transmission
2nd degree-intermittent complete blockage of transmission
3rd degree-continuous complete blockage of transmission
Disturbances of the origin of stimulation include:
rate of SA node erratic resulting in erratic heart beat
more than one natural pacemaker site
high heart rate, e.g., sinus tachycardia
low heart rate, e.g., sinus bradycardia
Pacemakers are indicated to improve cardiac output, prevent symptoms,
or protect against arrhythmias related to cardiac impulse formation and con-
duction disorders.

TYPES OF CARDIAC PACEMAKERS

A pacemaker has two physical parts, the pulse generator and the lead
system (Figure 21-1). The pulse generator produces electrical stimulation
pulses. Through the lead system, these pulses are delivered to the heart mus-
cle, causing the heart to contract. There are three types of pacemakers:
implantable, external invasive, and transcutaneous. For an implantable pace-
maker, both the pulse generator and the leads are placed inside the patient's
body without any exposed parts. For an external invasive pacemaker, the
pulse generator is located outside the patient's body, while the lead wire con-
necting the pulse generator and the heart muscle is inserted through a vein
into the right chamber of the heart. A transcutaneous pacemaker, sometimes
called an external noninvasive pacemaker, has a pair of skin electrodes
placed anterior and posterior to the chest. The electrical stimulation from the
 Biomedical Device Technology: Princ$les and Design

Pulse generator
Lead (insulated wire)
A

~lectrodetip

Figure 21-1. Implantable Pacemaker and Lead.

external pulse generator is conducted through the heart across the external-
ly placed electrodes.
According to how it regulates the pacing rate, a cardiac pacemaker has
three modes of operation: asynchronous, demand, and rate-modulated. A
pacemaker in asynchronous mode can deliver only a fvted rate of stimula-
tion. In demand mode, it senses the heart's activity to determine its pacing
sequence. Rather than pacing at a fixed rate, a rate-modulated pacemaker
can adjust its pacing rate based on the state of physical activity of the patient;
therefore it is able to adjust the patient's cardiac output to meet the body's
demand.
Some pacemakers can perform cardiac defibrillation. An implantable
cardiac defibrillator (ICD) automatically produces a shock to the heart upon
detecting ventricular fibrillation. Ventricular fibrillation is a deadly form of
arrhythmia caused by completely uncoordinated contraction of heart mus-
cle. Cardiac defibrillation is discussed in Chapter 22.
Most modern pacemakers are programmable. Parameters that can be
programmed include pacing rate, mode of pacing, pacing pulse amplitude,
pulse duration, sensitivity, et cetera. In addition to the simple programmable
features, multiprogrammable pacemakers have built-in programmable diag-
nostic tests as well as the ability to log heart rhythms and pacing activities.
To program or interrogate an implanted pacemaker, a pacemaker program-
mer or receiver is placed on the skin surface above the pacemaker.
Programming commands and data are transmitted via telemetry (or electro-
magnetic coupling) between the programmer and the pacemaker.
 Cardiac Pacemakers

PACEMAKER LEAD SYSTEM

The pacemaker lead system serves two functions. The first is to transmit
pacing pulses from the pulse generator to the heart. The second is to pick up
electrical activities of the heart to modify the pacing sequence. The pace-
maker lead is insulated with nonconductive material (e.g., silicon) except at
the tip electrode and the connector to the pacemaker. The conductor is made
of corrosion-resistive wire, which is coiled to increase its flexibility. The elec-
trode (tip of the lead wire) may be attached to the surface of the heart or
inserted through a vein into the chambers of the heart. The former is called
the myocardial (or epicardial) lead system and the latter is called the endo-
cardial lead system. Figure 21-2 shows the two lead systems.

Pacemaker lead Pulmonary

/
R~ghtventricle
(a) (b)

Figure 21-2. Pacemaker Lead System (a) Endocardial Lead, (b) Myocardial Lead.

To complete the conduction path, the current produced from the pulse
generator, after passing through the heart tissue, must return to the pulse gen-
erator. For a unipolar lead configuration, a single conductor lead is used. The
conductor in the pacemaker lead carries the pacing current from the pulse
generator circuit to the heart tissue. The metal housing of the pacemaker
serves as the return electrode. As there is only one conductor in the pace-
maker lead, the return current must therefore return via the conductive body
tissue of the patient to the metal housing and then back to the pulse genera-
tor circuit of the pacemaker. In a bipolar lead configuration, both the active
and return conductors are inside the insulated lead (a dual conductor lead).
The pacing current (from the electrode tip) to the heart is picked up by the
return electrode located near the tip electrode and returned to the pulse gen-
 Biomedical Device Technology: Principles and Design

Single conductor lead - &

-'
Dual conductor lead

~istributed Stimulation
return current current
(a)
Return
electrode
Electrode tip
Figure 21-3. Lead Configuration: (a) Unipolar Lead, (b) Bipolar Lead.

erator via the return conductor in the pacemaker lead wire. Figure 21-3 illus-
trates the two configurations.
A pacemaker with two leads (one in the atrium and the other in the ven-
tricle) to allow pacing and/or sensing of both the ventricle and the atrium is
referred to as a dual chamber pacemaker (Figure 21-4). A dual chamber
pacemaker can more likely restore the natural contraction sequence of the
heart.
In conventional cardiac pacing, only the right atrium and the right ven-
tricle are paced. During normal intrinsic heart contractions, both ventricles
are activated at almost the same time. Under conventional (right-heart-only)
cardiac pacing, contraction of the left ventricle is triggered through propaga-
tion of depolarization from the right ventricle. Such delay results in dimin-
ished cardiac output, which may cause significant problems for some
patients. Biventricular pacing refers to pacing of both ventricles simultane-
ously to improve cardiac output. This is achieved by placing an additional
lead in the lateral or posterolateral cardiac vein (located at the far side of the
left ventricle). Special lead placement techniques are required as it is not easy
to insert a lead into these cardiac veins.
 Cardiac Pacemakers

er leads

Pacemaker

Ventricular lead

Figure 21-4. Dual Chamber Pacemaker.

IMPLANTATION OF PACEMAKER

The procedures to implant myocardial and endocardia1pacemaker leads

are quite different. The following descriptions summarize the two proce-
dures.

Myocardial Pacemaker Lead

Myocardial pacemaker lead implantation is performed under general
anesthesia. The procedure starts with an incision between the ribs to expose
the apex of the heart. An area of the apex free from coronary arteries is cho-
sen. The electrode tip is inserted into the heart muscle (screw-in for spiral
electrode or stab-in for barb electrode) at the chosen location. The other end
of the lead is tunneled under the skin down toward the abdomen. A shallow
incision is made in the abdomen so that the lead can be pulled out. Correct
lead placement is confirmed by verifying the pacing and sensing thresholds
(often by using a pacemaker analyzer). The lead is then plugged into the
pulse generator. The pulse generator is pushed into the pocket made by the
incision. The two incisions are closed to complete the procedure.

Endocardia1 Pacemaker Lead

Endocardia1 pacemaker lead implantation is done under local anesthesia.
The procedure starts with an incision over a vein (e.g., the right external
jugular or subclavian). The pacemaker lead is passed down within the vein,
into the right atrium, through the right atrioventricular valve, and into the
right ventricle. Correct placement is achieved when the tip of the lead is
 Biomedical Device Technology: Principles and Design

wedged firmly between the trabeculae at the apex of the right ventricle and
the lead is observed to be fixed and immobile. Fluoroscopy is often used dur-
ing the procedure to ensure proper lead placement. A shallow incision is
then made in an appropriate area of the upper chest (e.g., under the clavicle).
A tunnel is formed under the skin of both incisions so that the connector end
of the lead wire is pulled through and is accessible at the chest incision.
Correct lead placement is confirmed by verifying the pacing and sensing
thresholds. The lead is then plugged into the pulse generator. The pulse gen-
erator is pushed into the pocket made by the incision. The two incisions are
closed to complete the procedure.
After the procedure, the pacemaker parameters are programmed accord-
ing to the patient's condition. Regular patient follow-up should be scheduled
to monitor the condition of the pacemaker's battery and to confirm that the
programmed parameter values are appropriate. Note that the pacing thresh-
old (minimum values to achieve pacing) will rise shortly after implantation
and eventually become stabilized after about 3 to 4 months.
Figure 12-5a shows the strength duration curve of heart stimulation. In
order to stimulate the heart, the stimulus must have its amplitude and dura-
tion above the curve. In the example shown, if the pulse amplitude is 1 V,
the pulse width of the stimulation must be larger than 0.7 ms. A pulse dura-
tion of 1 ms or longer will be chosen. Figure 21-5b shows the variation in
voltage threshold after lead implantation. In this example, the initial pacing
voltage amplitude should be set to over 3 V and subsequently reduced to
about 2 V after 2 months. A lower pacing amplitude and narrower pacing
pulse width will prolong the battery life.
Pacemakers are presterilized in the package. If sterility is compromised,
resterilization should be performed (e.g., using ethylene oxide below 60°C
and 103 kPa). Most manufacturers specify not to resterilize pacemakers more
than twice.

A A
d
g 3-
(D

(b)
...................,.......................................
,

I I I > I i I I >
1 2 3 4 Time (ms) 1 2 3 4 Time (month)
Figure 21-5. (a) Strength Duration Curve; (b) Change in Threshold After Implantation.
 Cardiac Pacemakers

PACING MODE SELECTION

The pacemaker modes are defined in the NBG Code. NBG stands for
the North American Society of Pacing and Electrophysiology (NASPE) and
the British Pacing and Electrophysiology Group (BPEG) Generic. It is a set
of codes specifying the modes of operation of implantable pacemakers. It is
intended for quick identification of the functionality of the pacemaker in case
a pacemaker patient requires intervention. It supersedes the older ICHD
(Intersociety Commission on Heart Disease) Code. Each letter of the five-let-
ter NBG code describes a specific type of operation. Table 21-1 describes the
codes.
Although there are five letters in the NBG Code, some pacemakers may
have only the first three or four letters imprinted on the pacemaker.

Table 21-1.
-
NBG Pacemaker Code
Position I 0-None
Chamber Paced V-Ventricle
A-Atrium
D-Dual chamber (ventricle and atrium)
S*-Single chamber (ventricle or atrium)
Position I1 0-None
Chamber Sensed V-Ventricle
A-Atrium
D-Dual chamber (ventricle and atrium)
S*-Single chamber (ventricle or atrium)
Position I11 0-None
Mode of Response T-Triggered
I-Inhibited
D-Dual (triggered and inhibited)
Position IV
Rate Modulation
Position V 0-None
Multi-site Pacing A-Atrium
V-Ventricle
D-Dual (ventricle and atrium)
-

*Note: Manufacturer's designation only.

 Biomedical Deuice Technology: Princ$les and Design

The following examples illustrate how to interpret the NBG pacing code.
An A 0 0 pacemaker will pace the atrium (1st letter-A) at a fixed rate (i.e.,
at every sensor-indicated interval) irrespective of the intrinsic rate of the
heart. Figure 21-6 shows the timing sequence of such a pacemaker. AP in the
diagram stands for atrial paced. The dotted line indicates the timer, which
keeps track of the pacing intervals. When this timer reaches zero, a pacing
pulse is generated.

Sensor-~ndicatedlnterval Sensor-~nd~cated
- -
~nterval ,
%
< i

Figure 21-6. A 0 0 Mode of Pacemaker Operation.

A WIpacemaker monitors ventricular contraction (2nd letter-V). It will

pace the ventricle (1st letter-V) if it cannot sense ventricular contraction
(intrinsic rate slower than sensor-indicated interval). If ventricular signal is
sensed, it will inhibit (3rd letter-I) the pacing action. VVI pacemakers are
often used to treat patient with second degree heart block. Figure 21-7 shows
the timing sequence of such a pacemaker. In the diagram, VP stands for ven-
tricle paced and VS ventricle sensed. After the first paced ventricular con-
traction, the timer started to count down. It was reset by the sensed intrinsic
ventricular contraction. When the timer has reached the sensor-indicated
interval without sensing any intrinsic ventricular activity, a pacing pulse is
generated to trigger ventricular contraction.
A DDD pacemaker senses both atrial and ventricular activities (2nd let-
ter-D). When it cannot detect atrial contraction (2nd letter-D), it will trigger
the atrium (1st letter-D, 3rd letter-D). After an atrial paced or intrinsic con-
traction, if no ventricular contraction is detected (2nd letter-D) after the AV
delay interval, the ventricle will be paced (1st letter-D). If ventricular con-
traction is detected, the ventricular stimulation will be inhibited (3rd letter-
D). A DDD pacemaker is a dual chamber pacemaker; that is, it has two
leads, one in the right atrial chamber and the other in the right ventricular
chamber. The timing sequence of a DDD pacemaker is shown in Figure
21-8. In the diagram, AP stands for atrial paced, AS for atrial sensed, VP for
ventricular paced, and VS for ventricular sensed. If it is also a rate-modulat-
 Cardiac Pacemakers

i Sensor-indicated interval

Figure 21-7. WI Mode of Pacemaker Operation.

ed pacemaker, the pacing rate (or the sensor-indicated interval) will change
with the patient's activities. Such a pacemaker will be labeled DDDR. The
fourth letter indicates that it is a rate responsive pacemaker.

Sensor-indicated
Sensor-indicated Sensor-indicated interval (AV delay)
: interval (atrial)
. . , . I .<-. .-.... .....
:.

Figure 21-8. Example of DDDR Mode of Operation.

Other than the modes described by the NBG Code, most pacemakers
can be switched to a magnet operation mode. Magnet mode is activated by
placing a magnetized programming head or a permanent magnet over the
pacemaker. During magnet operation, the pacemaker is paced asynchro-
nously at a predetermined fixed rate. Magnet mode is usually combined with
a threshold margin test and a self-diagnostic test to evaluate the integrity of
the lead and pacemaker system.
 Biomedical Device Technology: Principles and Design

PERFORMANCE CHARACTERISTICS

Table 21-2 lists some common implantable pacemaker performance

parameters and their nominal values.

Table 21-2.
Performance Parameters of an Implantable Pacemaker.
-
- - -- - -

Parameter
--
Capability
---- --
Nominal Setting
Lower rate 30 to 175 m i d (f2 m i d ) 60 min-'
ADL rate
Upper sensor rate
Amplitude
pulse width 0.12 to 1.5 ms ( f 2 5 ps)
Atrial sensitivity 0.25 to 4 mV (+40?0)
Ventricular sensitivity 5.6 to 11.2 mV (f40%))
Refractory period 150 to 500 ms (f9 ms)
Single chamber hysteresis 40 to 60 min-' ( f 1 m i d ) Off
Rate limit 200 min-I ( f 2 0 m i d ) Nonprogrammable

Lower rate is the programmed minimum pacing rate in the absence of

sensor-driven pacing. The time between two consecutive pulses at the lower
rate is called the escape interval.
Activities of daily living (ADL)rate is the sensor-driven target rate that the
patient's heart rate is expected to reach during moderate exercise.
Upper sensor rate is the upper limit of the sensor-indicated rate during
exercise.
Sensitivity is the voltage level to which the pacemaker's sense amplifiers
are responsive to electrical activities in the heart.
actor^ period is the period of time following the onset of an action
potential during which the heart tissue will respond to neither an intrinsic
impulse nor an extrinsic impulse.
Hysteresis is a pacing operation that allows a longer escape interval after
a sensed intrinsic event. By waiting a bit longer after an intrinsic impulse, it
gives the heart a greater opportunity to beat on its own.
Rate limit is a nonprogrammable upper limit of the pacemaker. It is a
built-in safeguard to limit the rate of the pulse generator.
In general, the acceptable lead impedance of the pacemaker is from about
200 to 1,000 LR. Very often, a patient load of 500 LR is chosen to verify pace-
maker parameters.
A number of different output waveform are used by different pacemaker
 Cardiac Pacemakers

manufacturers. See Figure 21-9 for different types of waveforms. The sim-
plest waveform is the square pulse.

Rectangular Trapezoidal Triangular Exponential

Figure 21-9. Pacemaker Output Pulses.

I‘c
Biphasic

A lithium-iodine battery is commonly used in implantable pacemakers.

For example, the Medtronic Sigma 213 lithium-iodine battery is rated at 2.8
V with a capacity of 0.83 Ah. Lithium silver vanadium (Li/SVO) batteries
with energy density of 2.0 kJ/cm3 are used in implantable cardiac defibrilla-
tors. In general, the lower the pacing rate, pulse amplitude, and pulse dura-
tion, the longer the battery life.
Chassis-the battery and electronic components of a pacemaker are
placed inside a polypropylene container and then totally shielded by a tita-
nium housing. The lead connectors are molded into an epoxy housing. A
radiopaque ID code is placed inside the housing so that the pacemaker can
be identified from a radiograph of the patient.

Example
Estimate the battery life of an implantable pacemaker given that it is pacing
100010 at a rate of 70 beats per minute, with pulse width of 0.45 ms, pulse
a.
amplitude of 3.5 V, and lead impedance = 500 The useful capacity of the
battery is 0.83 Ah at 2.8 V. Assume that 80% of the battery energy is used to
produce the impulses.

Solution
The average output power Pis the product of the output voltage, output cur-
rent, and the duty cycle, Therefore,
 Biomedical Device Technology: Princiljles and Design

The energy E of the battery is

E =V X I X t = 2.8 X 0.83 X 60 X 60 = 3.4 kJ
If 80% of the energy is used to produce the impulses, the longevity t of the
battery is
t=-= 3.4 lo' s = 0.21 x 10Y s = 6.6 years.
P 13 x lo-''

FUNCTIONAL BUILDING BLOCKS

OF AN IMPLANTABLE PACEMAKER

In an implantable pacemaker, the battery occupies most of the volume

of the pulse generator. The remaining space contains the electronic circuit for
sensing, pacing, and communication. Pacemakers are non-field serviceable;
all malfunctioned pacemakers are sent back to the factory for analysis. They
will not be repaired and reused. Figure 21-10 is a functional block diagram
of a dual chamber pacemaker. The descriptions of the various functional
blocks follow.
Battery Monitor-This is a voltage level detector to monitor battery
level. When the battery voltage drops below a certain limit (e.g., below
2.2 V for a 2.8 V lithium-iodine battery), the pacemaker will suspend
most activities (e.g., cease to collect data) and fall back to a fixed rate
to conserve power. A battery replacement message will be displayed
when the programmer is engaged with the pacemaker.
Programming-The transmitlreceive coil is the antenna for the pro-
gramming signal. The signal from the external programmer is detect-
ed and demodulated to provide input signal to program the pacemak-
er parameters. Stored data from the pacemaker can be modulated and
transmitted to the programmer via the transmitlreceive coil.
Atrial and Ventricular Sensing Amplifiers-The input filter and ampli-
fier function together provide bandpass and gain characteristics to
identify the intrinsic atrial and ventricular activities. The sensitivity
detector circuit senses a minimum amplitude level according to the
programmed sensitivity setting. A blanking circuit disables the sensing
circuit for a predetermined period of time after an impulse is deliv-
ered.
Output Circuit and Rate Limit-The output circuit is designed to deliv-
er an electrical pulse to the heart. The pulse amplitude and duration is
controlled by the digital timing circuit. The rate limit circuit operates
independent of other circuits in the pacemaker. It limits the pulse rate
 Cardiac Pacemakers

to a maximum value for patient safety (e.g., 185 min-' for ventricular
pacing).

Battery
power source
I 1 I

Battery
monitor >
- Reedswitch -
Programming
signal
mod/demod - >
v
Amplifier Sensibvity
Blanking 4 4 level
andfilter
detector
4
Ventricle Output
• circuit and
rate limit
<
t -0
case -
Output
circult and
rate limit
<
Atrium
\L/
Amplifier Sensitivity
Blanking and filter level

h
u detector

Clystal
oscillator >
Figure 21-1O.Functional Block Diagram of a Cardiac Pacemaker.

TEMPORARY PACING

Temporary cardiac pacing is achieved by employing external pacemak-

ers. Temporary pacing is used under the following situations:
After open-heart surgery until the heart reverts to a satisfactory condi-
tion
 Biomedical Device Technology: Princtjiles and Design

For temporary pacing until a permanent pacemaker can be implanted

When permanent pacing is not necessary such as in the case of some-
one recovering from myocardial infarction
During surgery to control the heart rate
There are two types of external pacemakers: invasive and noninvasive.
External inuasive pacemakers have the pulse generator located outside the
patient's body. The pacing lead is inserted into the heart chamber via a
venous access (endocardial lead). A return lead is attached to the ground
electrode placed on the patient's skin.
External noninvasivepacemakers (alsoknown as transcutaneous or transtho-
racic pacemaker) have two large-surface, pre-gelled, adhesive, disposable
electrodes that conduct the stimuli through the skin and skeletal muscle to
the heart. One electrode supplies a current and the other collects the current
from the body. They are applied to either the anterior-posterior or anterior-
anterior position of the patient's chest. External noninvasive pacing is pre-
ferred in some situations because it can be applied more quickly and easily
and does not require the skills of a physician. However, each stimulating
pulse causes contraction of skeletal muscle and can be painful to the patient.
In demand mode, intrinsic heartbeats are generally picked up by an inte-
grated ECG monitor using a separate set of leads and electrodes. As the elec-
trical current must overcome the impedance of the skin and underlying tis-
sues, and only a portion of the current flows through the heart, the pacing
energy delivered by a transcutaneous pacemaker is substantially higher than
that of an implantable or external invasive pacemaker (e.g., pacing current
up to 20 mA and pulse width up to 20 ms in transcutaneous pacing). Table
21-3 summarizes the differences of the three types of pacemakers.

Table 21-3.
--
Characteristics of Different mes of Pacemakers.
- -- - -- -

- .
- -
Implantable External Invasive Tranrnrtaneous-
-

Pulse Generator Inside body Outside body Outside body

Placement
Lead Placement Inside body; Inserted through Outside body;
connected to heart skin via a blood on skin posterior
via a blood vessel vessel into the and anterior to
into the heart heart chamber the chest
chamber
Pacing Energy Level Low Low High
- - - -- -
 Cardiac Pacemakers

POTENTIAL PROBLEMS WITH PACEMAKERS

Although the electronic circuits of modern pacemakers are very reliable,

the pacemaker/lead system may operate inappropriately due to the follow-
ing problems:
Threshold Drij-Threshold level may increase after initial implantation
and may drift higher after prolonged usage. Increase in threshold may result
in failure to capture.
Lead Problems-Broken lead wires, poor connections, and insulation fail-
ure may cause continuous or intermittent loss of capture, failure to sense
properly, loss of sensing, cross-talk between leads, and inhibition of pacing.
Cross-talk occurs in dual chamber pacemakers when a stimulus or intrin-
sic event from one chamber is sensed by the other chamber (e.g., the ven-
tricular lead senses the pacing stimulation initiated in the atrium), resulting
in an inappropriate pacemaker response such as inhibition or resetting of the
refractory period.
External Interference-Pacemakers may be susceptible to certain sources of
electromagnetic interference (EMI).They include, but are not limited to, the
following:
Magnetic resonance imaging (MRI) scanners
Therapeutic diathermy devices
Therapeutic ionization radiation equipment
Defibrillators
Electrosurgical/cautery units
High voltage systems and current-carrying conductors
Radio transmitters
Theft prevention security systems
High power electromagnetic fields
Ultrasound energy
Communication equipment, such as cellular phones
If the pacemaker is inhibited or has reverted to asynchronous operation
in the presence of EMI, turning off the source or moving the pacemaker
away from the source may return the pacemaker to normal operation.
Extremely strong EM1 sources, however, may reset the pacemaker to partial
or full electrical reset condition.
Muscle sensing-Electrical signals from skeletal muscle activities may be
sensed and misinterpreted as heart activities. Muscle sensing may be cor-
rected by increasing the sensitivity of the pacemaker.
Muscle stimulation-Refers to the stimulation of a muscle (other than the
heart) by a pacing stimulus. Skeletal muscle stimulation only occurs with
unipolar lead systems as it uses the body tissue as a return electrical path.
 Biomedical Device Technology: Principles and Design

Diaphragmatic stimulation can occur in either bipolar or unipolar systems,

usually due to electrode placement that is too close to the diaphragm or
phrenic nerve.
 Chapter 22

CARDIAC DEFIBRILLATORS

State the clinical applications of cardiac defibrillation. Sketch the

damped sinusoidal, truncated exponential, and biphasic defibrillation
waveforms, and analyze the basic circuits to generate such waveforms.
Draw a block diagram of a cardiac defibrillator and explain the func-
tions of each block.
Describe built-in safety features, including isolated output and energy
dump.
Identify and explain the functions of critical components in a typical,
defibrillator.
Explain synchronous cardioversion and its operating precautions.
Identify common problems of cardiac defibrillators and methods to
prevent the problems.

CHAPTER CONTENTS

1. Introduction
2. Principles of Defibrillation
3. Defibrillation Waveforms
4. Waveform Shaping Circuits
5. Functional Building Blocks of Defibrillators
6. Output Isolation and Energy Dumping
7. Cardioversion
8. Defibrillator Operation and Quality Assurance
9. Common Problems
 Biomedical Device Technology: Princ$les and Design

INTRODUCTION

Fibrillation is an arrhythmia. During fibrillation, the heart muscle quiv-

ers randomly and erratically as a result of individual groups of heart muscle
contracting randomly instead of synchronously. If fibrillation occurs at the
ventricles, it is called ventricular fibrillation. This situation is life-threatening
as it prevents effective pumping of blood to vital organs such as the brain,
lungs, and the heart itself. If it occurs at the atria, it is called atrial fibrillation.
Atrial fibrillation is less severe and in most cases is not fatal. However, it
compromises cardiac output and will likely lead to other more severe
arrhythmias. Figure 22-1 shows a normal sinus rhythm, atrial fibrillation,
and ventricular fibrillation waveforms.
Prevost and Batelli in 1899 proved that an "appropriate" large, alternat-
ing current (AC) or direct current (DC) electric shock could reverse ventric-
ular fibrillation. It was not until 1960 that open-chest defibrillation was
replaced by the external defibrillation method. Today, the defibrillator is a
critical life-saving medical device that is widely deployed in hospitals, clin-
ics, ambulances, and even in public areas to treat sudden cardiac arrest.

PRINCIPLES OF DEFIBRILLATION

Fibrillation can be caused by disruption of the electroconductive path-

ways in the myocardial muscle such as the SA or AV nodes. It may also be
triggered by an electrical shock. Passing a very large momentary electrical
current through the heart causes all musculature of the heart to be depolar-
ized for a short period of time and enter their refractory period together. This
gives the SA node a chance to regain control and return to normal rhythm.
Defibrillation can be external or internal. During defibrillation, an ECG
monitor is necessary to detect ventricular fibrillation and thereafter monitor
the heart function until the patient can be placed in a critical care environ-
ment. When defibrillation is applied externally on the patient, a larger volt-
age (and higher energy) is required to overcome the impedance of the body
and to allow enough current to go through the heart. A typical discharge
energy range is from 2 to 40 joules for internal defibrillation and 50 to 400
joules for external defibrillation.
Defibrillators/monitors can also be used for synchronized cardioversion
to treat certain atrial arrhythmias such as atrial flutter or atrial fibrillation. An
electrical shock applied to a nonfibrillating heart during the T wave of the
heart rhythm may trigger ventricular fibrillation. During atrial flutter or atri-
al fibrillation, the ventricles can still contract at regular intervals, therefore
 Cardiac Defibrillators

- -
Figure 22-1. (a)Normal Sinus Rhythm, (b)Atrial Fibrillation,
(e)VenMdw Fibrillatian.

producing a certain amount of cardiac output (e.g., 80% of the normal value).
It would be counterproductive and could create a life-threatening situation to
the patient should this countershock to correct atrial heart condition trigger
ventricular fibrillation. In order to avoid discharging the energy during a T
wave under cardioversion mode, the defibrillator is synchronized to dis-
charge its energy right after the patient's R wave and before the T wave. This
is achieved by detecting the R wave with the help of an ECG monitor and
electronically synchronizing the energy discharge with a short delay (e.g., 30
ms) from the R wave.
An implantable cardiac defibrillator (ICD) is a pacemaker with defibril-
lation capability. Upon sensing ventricular fibrillation, an ICD will automat-
 Biomedical Deuice Technology: Pfinc$les and Design

icaIly produce a shock to the heart. An ICD can be programmed to provide

defibrillation, cardioversion, antitachycardia pacing, and antibradycardia
pacing.
Studies since its initial conception indicated that numerous factors can
affect the effectiveness of the defibrillation procedure. These include the
waveform, energy and amplitude of the electric shock, the electrode position,
and interface impedance, as well as the size and weight of the patient.

DEFIBRILLATION WAVEFORMS

Early experimental defibrillators used 60 Hz alternating current (AC)

and a step-up transformer to create and increase the defibrillation voltage.
Bursts of several hundred volts of sine wave were applied across the chest
wall for a period of 0.25 to 1 second. The desire for portability led to the
development of direct current (DC) defibrillators. A D C defibrillator uses a
battery as the power source so that connection to the AC outlet is not
required during defibrillation. The battery may be replaced or. recharged
after use. It was later discovered that D C shocks were more effective than
AC shocks. Until recently, defibrillators have used one of the two types of
monophasic waveforms: damped sinusoidal (MDS) and monophasic trun-
cated exponential (MTE).The MDS waveform is also called the Lown wave-
form. Figure 22-2 shows a typical MDS and MTE defibrillation waveform.
Note that the typical MDS waveform has a small negative component; there-
fore, strictly speaking, it is not truly monophasic.
Monophasic waveforms require a high energy level (up to 360J) to defib-
rillate effectively. A MDS waveform requires a high peak voltage (e.g., 5,000
V) to deliver such energy. The MTE waveform uses similar energy settings.
However, it uses a lower voltage than the MDS waveform. In order to deliv-
er the same amount of energy, the MTE waveform requires a longer dura-
tion. While studies had associated myocardium damages with high peak volt-
ages, long-duration shocks have higher chances of refibrillation.
Studies in the early 1990s had shown that biphasic defibrillation wave-
forms are more effective than monophasic waveforms. In fact, the biphasic
waveform has been the standard waveform for an implantable cardiac defib-
rillator (ICD) since it was introduced. With biphasic waveforms, the defibril-
lation current passes through the heart in one direction and then in the
reverse direction. A number of biphasic waveforms are incorporated by dif-
ferent defibrillator manufacturers. Figure 22-3 shows one such waveform.
Studies have shown that defibrillations using biphasic waveforms not only
defibrillate as well as traditional monophasic waveforms but also are associ-
ated with better postshock cardiac function, fewer postshock arrhythmias,
 Cardiac Defibrillators

Time
f5 m s 3 <
- 20-40 ms >
-.
Damped sinusoidal Truncated exponential
Figure 22-2. Monophasic Defibrillation Waveforms.

and better neurological outcomes for survivors. In addition, biphasic defib-

rillators at lower energy settings have been shown to produce the same
results as traditional high energy monophasic defibrillators. One manufac-
turer recommends that escalating energy shock protocols traditionally used
in monophasic defibrillation are not required in monophasic defibrillation.
Instead, a setting of 150J is recommended to be used for the first and all sub-
sequent shocks.

< 5-20ms >-

Figure 22-3. Biphasic Defibrillation Waveforms.
 Biomedical Device Technology: Principles and Design

WAVEFORM SHAPING CIRCUITS

Figure 22-4 shows a simple functional block diagram of a D C cardiac

defibrillator. The main component of a defibrillator is the energy storage
capacitor. The capacitor is charged by the charging circuit, which is powered
from the power supply. The charge control circuit monitors the amount of
energy stored in the capacitor. It terminates the charging process when ade-
quate energy has accumulated. The discharge control releases the stored
energy from the energy storage capacitor when the user activates the dis-
charge buttons. The waveform shaping circuit produces the particular type
of waveform to effect defibrillation. Three of the more common waveform
shaping circuits are discussed in this section.

AC power - Power Charging Energy Waveform To patient

source supply circuit storage shaplng Lb paddle

User input
F] I
Discharge

I
Figure 22-4. Block Diagram of a DC Defibrillator.

Monophasic Damped Sinusoidal Waveform

Figure 22-5 shows a simplified circuit to produce a damped sinusoidal
defibrillation waveform. It consists of the energy storage capacity C, a step-
up transformer, a rectifier, a charge relay, a discharge relay, and a wave shap-
ing inductor L. During charging, the charge relay is energized, and AC volt-
age from the power source is stepped up to the desired level by the step-up
transformer. The charging circuit (a full wave rectifier in this example) con-
verts AC to a D C charging voltage. The capacity Cis being charged up by
this high-level DC voltage until enough energy is stored in the capacitor. The
energy Ec stored in the capacitor is related to the voltage across the capaci-
tor according to the equation:
 Cardiac DeJibrillators

The voltage across the capacitor is monitored to determine the amount

of energy stored. The charge relay is deenergized when sufficient energy is
stored in the capacitor. When the operator pushes the discharge buttons, the
discharge relay is energized. The energy stored in the capacitor flows
through the inductor L into the patient. For ease of analysis, the patient load
R is considered to be a resistive load of 50 R. The current discharging
through this LRC circuit produces the damped sinusoidal waveform (Figure
22-2a).

A cardiac defibrillator is designed to deliver up to 400 joules of energy dur-

ing discharge. If a capacitor of 16 pF is used as the energy storage capacitor,
what is the minimum voltage across the capacitor at full charge?

Solution
Using equation (I),

The minimum voltage across the capacitor is 7,000 V.

In practice, not all energy stored in the capacitor is delivered to the

patient during discharge. Some energy, for example, is lost as heat in the dis-
charge circuit. The energy delivered to the patient EDis always less than the
energy stored in the capacitor Ec. Assuming that the patient load is a con-
stant resistive load, the energy delivered to the patient EDfor an MDS wave-
form defibrillator is given by:

ED =I t='o

t=o
V2
d t = -
R R
It="v2dt.
t=o

A typical value of the waveform shaping inductor to produce an MDS

waveform is 50 mH. The function of the resistor RL is to limit the initial
inrush current into the capacitor when the charge relay is first energized.
Without RL, the large inrush current may damage components in the charg-
ing circuit. A typical value of RL is 3 kfl, which will limit the initial inrush
current to a worst case of 2.3 A (7,000 V divided by 3 KR).
 Biomedical Device Technology: Princ$les and Design

-----J

Patient load

4-
I
Step-up
transformer

Charge relay Discharge relay

Figure 22-5. Simple MDS Defibrillator Circuit.

Monophasic Truncated Exponential Waveform

Figure 22-6 shows a simplified circuit of a monophasic truncated expo-
nential waveform defibrillator. This circuit is identical to the circuit described
above except that there is no waveform shaping inductor in the discharge cir-
cuit. A typical value of the energy storage capacitor is 200 pF. Without the
inductor, the discharge circuit is an RC instead of a LRC circuit where R is
the patient load. An RC discharge will produce an exponential decay curve
(Figure 22-2b). Instead of allowing sufficient time to discharge all energy
stored in the energy storage capacitor, a MTE defibrillator will terminate the
discharge when enough energy is delivered to the patient. During defibrilla-
tion, the voltage across the paddles is monitored and the amount of energy
discharged into the patient EDis determined by:

Biphasic Truncated Exponential Waveform

Figure 22-7 shows a simplified circuit of a biphasic truncated exponen-
tial waveform generator. A bank of switches is added to the previously
described MTE circuit. By closing and opening the biphasic switches S1 to
S4, a biphasic waveform (Figure 22-3) is produced. Table 22-1 shows the
switching sequence of the switches and relays for the charging, discharging,
and energy dumping functions. In the table, an "X" denotes switch closure.
 4-

-+
From
I

Charge relay

chargrng
circuit

2.-

3
SC
I

I
I

j
I
I

I
I
I

Charge relay
Stepup
transformer

-
a

--
-- C

a
+
R
Cardiac Dejibrillators

+I
I!
Discharge relay
Figure 22-6. Simple MTE Defibrillator Circuit.

There are four phases in the discharge sequence: positive (Pl),zero (P2),neg--
ative (P3), and discharge (P4).
During the charging period, the charge relay is energized, and switch SC
is closed so that the capacitor (e.g., a 200 pF metalized polypropylene capac-
itor) is charged by the charging circuit. SC will open when enough charge is
stored in the capacitor. In the positive phase of the discharge sequence, S1,
S4, and SD are closed. The flow path of the current from the capacitor is:

-
s3

s4
1'
+jL+
I
T 3
i
-----!
SP
X! *-@--- - -,
jI
I
I
I
I
I
I
I
I

I
2

Drscharge relay
Figure 22-7. Simple BTE Defibrillator Circuit.
0;+;

Patrent load
 Biomedical Device Technology: Principles and Design

R+Sl+SD-l+patient+SD-2+S4 and back to the capacitor. During the

zero phase, only SD is closed. This zero phase provides a time separation to
ensure that Sl and S4 are opened before S2 and S3 are closed. If all fours
switches are closed at the same time, a short circuit on the output circuit will
result. In the negative phase, S2 and S3 are closed. The current flow path
from the capacitor is: R+S3+SD-2+patient+SD-l+S2 and back to the
capacitor. After enough energy is discharged, all biphasic switches and SD
are closed to remove the remaining charge stored in the energy storage
capacitor. The previously described phases complete the discharge sequence.
If the capacitor is charged but defibrillation is not necessary, the energy
stored in the capacitor must be removed for safety reasons. All modern
defibrillators have a programmed dumping sequence to remove the stored
charge if defibrillation is not performed within a set time (e.g., after 60 sec-
onds) and also when another energy level is selected (to avoid high residue
energy stored in the capacitor). To dump the stored energy, all biphasic
switches are closed to allow the energy to discharge through the resistor R.

Table 22-1.
Biphasic Waveform Generator Switching Sequence.
- - - - -- - - -

SC $7 S2 S3 S4 SD - -

Charging X
P1 (1-2) X X X Positive
P2 (2-3) X Zero
P3 (3-4) X X X Negative
P4 (4-5) X X X X X Discharge
-
Energy dump
- -
X X
-
X X - - -- - - -

SC and SD are mechanical switches. In fact, these are the contacts of

high current relays (or contactors). The biphasic switches Sl, S2, S3, and S4
are solid-state switches. S4 is usually a high-voltage, high-current switch (such
as an insulated gate bipolar transistor) that is designed to interrupt the circuit
at any current level. S1, S2, and S3 are usually silicon controlled rectifiers.
The function of the resistor R is to limit current during capacitor discharge
or energy dump. Some manufacturers may introduce an additional inductor
m o d i f y the shape of the waveform. For example, one manufacturer used
a5 a, 700 pH inductive resistor to serve both functions.
 Cardiac Dejbrillators

FUNCTIONAL BUILDING BLOCKS OF DEFIBRILLATORS

Figure 22-8 shows the functional block diagram of a DC defibrillator.

After the user selects the energy setting and pushes the charge button, the
charge control circuit energizes the charge relay. The voltage across the ener-
gy storage capacitor is monitored during charging. Using equation (I), the
charge relay is deenergized when the voltage across the capacitor is equal to
the voltage corresponding to the selected defibrillation energy level. The
charging sequence is then completed. The discharge relay is energized once
the user pushes the discharge buttons on the defibrillator paddles (both but-
tons, one on each paddle, must be activated to prevent inadvertent dis-
charge). The energy stored in the capacitor is then released through the
waveform shaping circuit to the patient's chest to perform defibrillation. The
energy being delivered to the patient is determined by the voltage and cur-
rent monitors (ED= V x I x t). When the total delivered energy has reached
the user selected value, the discharge relay is deenergized.

Power
shapng + rnonltor < I

I
I
I
I
To
+pabent
padde

AC power
Charge Discharge

-
source
relay relay

t
User input
rF*--v-77-7
v
A

: $her@ .
I - 1
Charge Voltage Energy
j :annmapdy
> I

control monitor rnon~tor

,. n

t
I 1
I. *' %,'..- 'J
, P -
I a *
4
I ,
. * u s
I

, " . ',
It'* * - # + . ~ d

D~scharge
: :Enei&:-,, Energy + control -
: """
I , setting
,
I' 4, I
G r r v . r 1
-'
C~@fi98e
r.:c~a+i
L->--a >---I

Figure 22-8. Functional Block Diagram of a D C Defibrillator.

Due to portability requirements, an internal rechargeable battery is used

as the primary energy source of DC defibrillators. The capacity of a fully
charged battery is usually sufficient to perform 20 to 80 defibrillation dis-
charges. Defibrillators are always plugged into the AC mains on standby. AC
voltage is rectified to charge the battery. When fully charged, the battery
 Biomedical Device Technology: Principles and Design

charge is maintained using a trickle charge system. During the charging

phase, the low-voltage DC from the battery is first converted to a high-fre-
quency (e.g., 25 kHz) AC voltage by an inverter. This AC voltage is then
stepped up to a higher voltage, say, 7,000 V, and rectified to charge the ener-
gy storage capacitor. The functional block diagram of the power supply and
charging circuit is shown in Figure 22-9.

Low-voltage
high-frequency AC

From
AC
mains

Figure 22-9. Defibrillator Power Supply and Charging Circuit.

OUTPUT ISOLATION AND ENERGY DUMPING

A defibrillator produces an electrical shock on the patient to correct heart

arrhythmia. If during the delivery of an electrical shock, the operator inad-
vertently touches the discharge paddles or the patient, the shock may cause
bums or trigger ventricular fibrillation to the operator. Such injuries to the
operator can be prevented by isolating the output of the defibrillator from
ground. Figure 22-10a shows a nonisolated defibrillator output circuit. When
an operator who has a ground connection is in contact with the output while
the energy is discharge$ a current will flow through the operator to ground
and return to the energy storage capacitor. Figure 22-lob shows an isolated
defibrillator output circuit. During shock delivery, the energy storage capac-
itor is not connected to ground. Theoretically, even when the operator is
touching a defibrillator paddle, no current will flow through the operator as
there is no return path for such current. In practice, however, a small current
still flows through the operator.
Isolated output can also prevent secondary burn to the patient. For a
grounded output circuit, a secondary burn occurs when the patient is in con-
 Cardiac Dejbrillators

tact with ground; such ground connection will provide a alternative return
path for the discharge current. It will cause a burn at the ground contact site
if sufficient current flows through this patient ground path.
After the energy storage capacitor is charged, if no defibrillation is nec-
essary after a period of time, the charge in the capacitor will be dumped
through a high power resistor. This is a safety feature to ensure that no haz-
ardous high voltage is present in the unit for the safety of the users. In addi-
tion, if another energy level is selected, the charge stored in the capacitor will
be released before it receives its new charge. This is to prevent accumulation
of charge in the capacitor, especially if the new selection is of lower energy.

Charge relay Discharge relay

contact, contact,

Defibrillator
charging
circuit II
\

-
-
-
f. -...-.--
Current flow

Relav contact at

discharge position

-
-

-
-
/ Figure 22-10. (a) Nonisolated Output, (b) Isolated Output
 Biomedical Deuice Technology: Principles and Design

CARDIOVERSION

Cardiac defibrillators are useful in correcting ventricular fibrillation.

However, the defibrillation shock may trigger ventricular fibrillation in a
healthy heart. During atrial fibrillation or atrial flutter, only the atrial muscle
is contracting erratically. Studies have shown that a defibrillation pulse
applied during the refractory period (T wave) of the ventricles may induce
other more severe arrhythmias such as ventricular fibrillation. Therefore, a
synchronization circuit is required in cardioversion to avoid such complica-
tions.
The window of discharge for safe cardioversion is immediately after the
QRS complex and before the T wave. A cardioversion synchronizing circuit
consists of an R wave detector and a time delay circuit to synchronize the dis-
charge within this safety window (Figure 22-11). An enabling signal is sent
to the discharge control about 30 ms after detection of the R wave. When the
synchronous cardioversion feature is selected, care should be taken to check
whether the machine is able to lock onto the R wave (usually, successful
detection of the R wave is highlighted on the ECG display). The user may
have to increase the ECG sensitivity level to provide sufficient R wave
amplitudes for the R wave detector to lock onto the signal.

Patlent's ECG R wave Time To defibrillator

waveform detector b delay discharge control

Figure 22-1 1. Cardioversion Synchronous Module.

DEFIBRILLATOR OPERATION AND QUALITY ASSURANCE

The procedures for safe operation of a cardiac defibrillator are:

1. Apply electrolyte gel to the defibrillator paddles (for better patient
electrical contact and to reduce risk of burn by lowering the skin
resistance).
%et energy level and press the charge button.
3, Allow capacitor to charge until a ready signal is given.
4. Press the paddles against the patient's chest.
5 . Clear the patient area.
6. Press the defibrillator discharge buttons.
7. Check the patient's ECG waveform.
 Cardiac Dejbrillators

8. Repeat the above procedures if no sinus rhythm is detected (usually

with a higher energy setting).
Because a defibrillator delivers a high-voltage therapeutic pulse to criti-
cally ill patients, reliability of the device is crucial. A defibrillator should be
tested regularly to ensure its performance. Most hospitals require daily user
testing. Depending on the hospital's protocol, testing includes functional
checks and may include checking for correct energy delivery. Often, exten-
sive performance verifications, including battery capacity and defibrillation
energy accuracy, are done periodically (e.g., every 3 months) by biomedical
engineering personnel. When not in use, a defibrillator should always be
plugged into the AC wall outlet to ensure that the internal battery is charged
and ready for use.

COMMON PROBLEMS

Problems with defibrillators can be divided into two groups: hardware

problems and operational problems.

Hardware Problems
Batteries are considered as high-maintenance components in a defib-
rillator. Failure of batteries prevents successful defibrillation, causing
death. Common battery problems include battery not fully charged,
battery failure, and cell memory failure in some batteries. Common
batteries used in defibrillator are nickel-cadmium (NiCad), sealed lead
acid, and nickel metal hydride (NiMH).
Electronic components in general are quite reliable.
Due to the need to deliver high-energy discharge pulses, relay failure
is not uncommon in defibrillators. Relay contacts (especially the dis-
charge relay) may be pitted from arcing which creates high resistance
at the contact; or fused due to excessive heat from high discharge cur-
rent.
Common problems with the energy storage capacitor are excessive
leakage (which prevents the capacitor from maintaining the energy
level), and short circuit due to insulation breakdown.
Hardware problems can be prevented by periodic performance assur-
ance inspections and battery analysis.
 Biomedical Device Technology:Princ$les and Design

Operational Problems
Users not familiar with the operation of the defibrillator
Incorrect application of conductive gel, causing high paddle-skin resis-
tance; high current density; or current shunt path, which may lead to
unsuccessful defibrillation or patient injuries
Incorrect paddle placement resulting in current not passing through
the heart
Electrical shock to staff from gel spill, staff touching patient, or staff
touching paddles
Most user errors can be prevented by proper in-service training, period-
ic practice, and equipment standardization.

Problem Unique to Synchronous Cardioversion

Unit not picking up R wave properly (missing or too low R wave level
due to poor skin contact or too low sensitivity setting)
To prevent problems or hazards, the following must be done by the users
regularly:
Receive proper user in-service training.
B
Perform operational check by chargin and discharging into dummy
load (e.g., weekly).
Plug unit into wall outlet to maintain battery charge when not in use.
Clean unit, especially paddles, after every use to prevent dried gel and
dirt from building up.
Check quantities and expiration dates of all supplies (e.g., conductive
gel, ECG electrodes) that are with the unit.
Send unit periodically to biomedical engineering for complete func-
tional and calibration check.
 Chapter 23

INFUSION DEVICES

State the applications of IV infusion.

Describe the setup and identlfy the components of a typical gravity flow
manual infusion system.
List the common problems encountered in manual gravity flow infu-
sion.
Differentiate between infusion pumps and infusion controllers.
Analyze the pumping mechanisms of common infusion pumps.
Evaluate the safety and nvenient features of modern infusion pumps.
2
Draw a functional bl c diagram of an infusion pump and describe its
operation.
Review performance verification procedures of infusion pumps.
Identify factors affecting the accuracy of infusion pumps.

CHAPTER CONTENTS

1. Introduction
2. Purpose of IV Infusion
3. Types of Infusion Devices
4. Manual Gravity Flow Infusion
5. Infusion Controllers
6. Infusion Pumps and Pumping Mechanisms
7. Common Features
8. Functional Block Diagram
9. Performance Evaluation
10. Factors Affecting Flow Accuracy
371
 Biomedical Device Technology: Princ$les and Design

Infusion devices are used to administer fluid into the body either through
intravenous (IV) or epidural routes. Infusion devices for IV administration
are commonly referred to as IV devices. As the venous pressure is below 50
mmHg (about 0.6 mHzO), a 1-meter water column is sufficient to allow grav-
ity to overcome the venous blood pressure and drive the solution into the
blood vessel. Manual gravity flow IV infusion is used extensively in health
care facilities for general-purpose infusion. To allow more controlled and
accurate fluid delivery, more sophisticated devices have been developed. A
number of infusion devices are available for different applications. This
chapter studies the principles and applications of a few of these devices.

PURPOSE OF IV INFUSION

In general, four types of solutions are administered intravenously.

1. Water-Usually in the form of saline or dextrose, to prevent patient
dehydration.
2. Medications and electrolytes-lV administration of drugs and elec-
trolytes produces precise and fast-acting effects as it sends the drug
directly into the bloodstream without going through the process of
digestion and absorpti&xamples include IV cardiovascular drugs,
chemotherapy drugs, et cetera.
3. Nutrition-Although parenteral nutrition can be delivered through
enteral feeding, total or partial parenteral nutrition is administered by
IV infusion to patients when their normal diet cannot be ingested,
absorbed, or tolerated for a significant period of time.
4. Blood-Blood infusion may be performed by an IV infusion device.
However, some may require special infusion sets to avoid problems
associated with the relatively high viscosity of blood and potential
hemolysis of blood cells.
IV infusion is a procedure commonly performed in health care facilities.
Infusion devices can be found in most parts of a hospital, including the emer-
gency department, medical imaging areas, operating rooms, and so forth.
 Infision Devices

TYPES OF INFUSION DEVICES

Many different types of infusion devices are available in the market.

Each has its own characteristics and serves some special applications. In gen-
eral, infusion devices can be divided into two main groups: gravity flow infu-
sion devices and infusion pumps. A gravity flow infusion device relies on the
gravitational force exerted by a liquid column to push the fluid via a venous
access into the patient's bloodstream, whereas an infusion pump has a motor-
ized pumping mechanism to generate the positive pressure. Within the grav-
itation group are the manual gravity flow sets and the infusion controllers.
There are two types of pumps in the infusion pump group: volumetric and
syringe. Within the volumetric pump group are three different pumping
mechanisms: piston cylinder, diaphragm, and peristaltic pumping mecha-
nisms. Figure 23-1 shows the different types of infusion devices.

I Gravity flow devices

Volurnetnc pump Syrlnge pump

Figure 23-1. Types of Infusion Devices.

MANUAL GRAVITY FLOW INFUSION

The simplest infusion device is the manual gravity flow infusion set.
Figure 23-2 shows a typical gravity flow infusion set. It consists of a long
flexible PVC tubing with a solution bag spike at one end and a luer lock con-
nector at the other end. The following sections describe the functional com-
ponents of a gravity flow infusion setup.
 Biomedical Device Echnolo~:Principles and Design

Solution
bag spike +
-
Dr~p
chamber- ,-

Regulat~ng Y-injection site for

clamp p~ggybackinfus~on

Figure 23-2. Gravity Flow Intravenous Infusion Set.

IV Solution Bag
The solution bag contains the IV solution and comes in different sizes
(e.g., 500 cc, 1 liter, etc.). The bag is usually hung on an IV pole about 1.5 m
above the infusion site to create enough pressure to overcome the venous
 Infision Devices

pressure to cause infusion. Solution-filled glass bottles instead of disposable

bags are used in some developing countries.

Solution Bag Spike

The solution bag spike is a sharp-ended tubing connecting the set to the
IV solution bag. This sharp spike is pushed through the seal of the solution
bag to allow solution to flow from the bag into the line.

Drip Chamber
The drip chamber is a clear compartment that permits the clinician to see
the solution drops coming down from the solution bag. The size of the drop
nozzle is designed so that each drop of solution is 1/20 ml (or 1/60 ml for
slow flow rate sets). By counting the number of drops within a known time
interval, a nurse can calculate the volume flow rate of the infusion.

Regulating Clamp
The regulating clamp is used to control the volume flow rate of infusion.
It is also known as a roller clamp. By squeezing the roller over the flexible
PVC tubing, it changes the cross-sectional area of the lumen, thereby con-
t r ~ l l i infusion
~ e flow rate.

Y-injection Site
The Y-injection site provides a point of access into the infusion line.
Drugs or other solutions can be injected into the infusion fluid by punctur-
ing the injection port with a needle. To infuse a second solution when an infu-
sion line has already been established (e.g., medication, blood plasma, etc.),
a setup called piggyback infusion is used (Figure 23-3a). In this setup, since
the secondary solution bag is located at a higher level than the primary solu-
tion bag, only the solution from the second bag will flow downstream
through the Y-injection site. Flow of the primary solution will resume auto-
matically when the secondary solution bag becomes emptied.

Occlusion Clamp
An occlusion clamp is used to totally occlude or shut down the infusion
flow. Unlike the roller clamp, an occlusion clamp either fully opens the infu-
 Biomedical Device Ethnology: Principles and Design

Secondary infusion
IV line

Occlusion

Luer lock connector

L f

Figure 23-3. (a) Piggyback Infusion Setup, (b) Occlusion Clamp.

sion line or totally occludes the line. It is constructed from a piece of thick
plaskwith the infusion line threaded through a keyhole-shaped opening in
the middle (Figure 23-3b). The line is fully open when the PVC tubing is at
the larger opening of the keyhole. If the line is pushed to the narrow end, the
clamp will occlude the tubing and shut off the flow.

Luer Lock Connector

A luer lock connector is a special twist lock mechanism to ensure a secure
connection. To set up an IV infusion, a catheter is inserted into a vein; the
other end of the catheter is a male luer lock connector. After the IV line is
primed, it is connected to the catheter using the luer lock connector.
The procedure to prime the IV line is:
Suspend the IV solution bag on the IV pole.
Open the bag containing the IV line following sterile supplies han-
dling procedure.
Insert the solution bag spike of the line into the IV bag.
 Infision Devices

Open the roller clamp and the occlusion clamp to allow the solution
to flush all the air from the line.
Remove air bubbles trapped in the Y-injection site by inverting and
gently tapping it with a finger.
Close the roller clamp and connect the luer lock at the end of the line
to the luer lock at the catheter.
Squeeze and release the drip chamber compartment to fill about one-
third of the chamber with the IV solution.
Slowly open the roller clamp to set up the desired solution flow rate
(by counting the drops using a stopwatch).

Example
A nurse is observing the drop rate in the drip chamber to set the infusion
flow rate on a manual gravity flow infusion set. How many drops per minute
should be counted in the drip chamber if an infusion flow rate of 60 ml/hr is
required? Assume that a 20 drops/ml nozzle is used in the drip chamber.

Solution
At a flow of 60 ml/hr, 60 X 20 drops will come down from the nozzle in 1
hour. Therefore, there will be 60 x 20/60 = 20 drops from the nozzle in 1
minute.

the fluid to be infused flows directly into the bloodstream, infusion

sets a e sterilized inside their packages and are single-use disposable devices.
With proper handling, infusion fluid from the solution bag flows only inside
the infusion line, thereby maintaining the sterility of the system. A major
drawback of manual gravity flow infusion is the low flow rate accuracy. As
the mechanism to regulate the infusion flow rate (roller clamp) relies on com-
pressing the PVC tubing of the infusion set, the rate of infusion cannot be
precisely controlled. Figure 23-4 shows the change in flow rate after initial
setup when there is no user intervention. This decrease in flow is due pri-
marily to the nonperfect elastic nature of the material of the infusion line.
The flow rate also changes with the level of fluid in the solution bag (the
height of the liquid column). Experiments have shown that the flow rate can
drop by about 40% within a couple of hours after initial setup. To overcome
such a problem, as a normal practice, a nurse must recheck the flow rate after
initial setup and adjust the regulating clamp to reestablish the desired flow
rate.
 Biomedical Device Technology: A.inc$Zes and Design

I
I I b
I 2 3 Time (hours)
Figure 23-4. Flow Rate Change in Gravity Infusion.

INFUSION CONTROLLERS

An infusion controller overcomes the problem of flow rate variation by

automatically adjusting the regulating clamp. Figure 23-5 shows the setup of
an infusion controller. An infusion controller monitors the flow rate by
counting the drops in the drip chamber. A typical drop sensor consists of an
infrared light-emitting diode (LED) and an infrared light-sensitive transistor,
each located on the opposite side of the drip chamber. A fluid drop from the
solution bag interrupts the optical path and produces an electrical pulse. The
flow rate is computed from the drop rate and the drop size. The calculated
flow rate is then compared to the setting. If it is lower than the setting, the
orce of the pinch mechanism will be released to allow more fluid
pinchin?
to flow hrough. If it is higher, it will increase the pinching force to reduce
the flow. Such a feedback mechanism maintains a constant flow rate equal to
the setting. Although it automatically regulates the fluid flow rate, an infusion
controller still relies on gravitational force to generate the infusion. If there is
some restriction in the infusion line, the gravity pressure created by the liq-
uid column may not be sufficient to produce the desired flow rate.

INFUSION PUMPS AND PUMPING MECHANISMS

An infusion pump contains a motor-driven pumping mechanism to pro-

duce a net positive pressure on the fluid inside the infusion line. With the
 Injision Devices

Drop
sensor

Figure 23-5. Infusion Controller.

pumping m
ith anism, infusion pumps produce a more controlled and con-
sistent flow an infusion controllers. Infusion pumps can be divided into two
types: volumetric pumps and syringe pumps.
Three common pumping mechanisms are used in volumetric infusion
pumps. They are piston cylinder, diaphragm, and peristaltic. A syringe
pump uses a screw and nut mechanism to drive the plunger of a syringe; it
is also called a screw pump. The following sections describe these pumping
mechanisms.

Piston Cylinder Pumps

The pumping mechanism of a piston cylinder pump consists of a cylin-
der, a piston, and valves that are mechanically linked to the piston motion.
A stepper motor drives a cam to move the piston in and out of the cylinder
in a reciprocal motion. Figure 23-6a shows that when the piston is moving
 Biomedical Device Technology:Pfinc$les and Design

downward, it creates a negative pressure inside the cylinder. The valve,

which is linked to the cam, will be in such a position that the input port to
the cylinder is opened and the output port is closed. Fluid from the IV bag
will therefore be drawn into the cylinder. When the piston is moving upward
(Figure 23-6b), the valve will close the input port and open the output port,
allowing IV solution in the cylinder to exit through the output port. The
stroke distance and the diameter of the piston determine the stroke volume,
and the infusion flow rate is equal to the stroke volume times the cam's rota-
tional speed.

, Flow control valve

Direction of
fluid flow
Figure 23-6. Piston Cylinder Infusion Mechanism.

Diaphragm Pumps
1
The pu ping mechanism of a diaphragm pump is similar to that of a pis-
ton cylinder pump except that the stroke motion is replaced by a moveable
diaphragm. In the illustration shown (Figure 23-7), when the diaphragm
moves to the left, the intake valve is open to allow fluid to enter the fluid
chamber. When it moves to the right, fluid is forced out of the chamber.
Repeating the action provides a continuous flow of fluid.

Peristaltic Pumps
A peristaltic pump employs a protruding finger mechanism to occlude
the flexible IV tubing. Its pumping action is similar to one using the thumb
and index finger to squeeze on a plastic tubing filled with fluid and then run-
ning the fingers along the tube. This action will force the fluid to move along
 Infision Devices

Fluid-filled
chamber

'L
One-way valves
Diaphragm Direction of
fluid flow

Figure 23-7. Piston Diaphragm Infusion Mechanism.

the direction of the finger motion. Repeating this action will produce a con-
tinuous fluid flow.
Figure 23-8a shows the pumping mechanism of a rotary peristaltic infu-
sion pump. In a rotary peristaltic pump (or roller pump), the rotor has sev-

P
eral protru ing rollers. The flexible IV tubing is placed inside a groove on
the pumpi g mechanism housing with one side open to the rotor. The rollers
on the rotating rotor push the tubing against the wall of the groove. The pro-
truding rollers, while occluding the tubing, move in one direction along the
IV tubing, creating a continuous fluid flow in the direction of motion of the
rollers.
Instead of rotating the protruding rollers over the IV tubing, the pro-
truding fingers in the linear peristaltic infusion pump sequentially occlude
the IV tubing. Figure 23-813 shows the positions of the protruding fingers of
a linear peristaltic pump at three sequential time instances. These coordinat-
ed motions of the protruding fingers produce a continuous flow of fluid in
the direction shown. The driving mechanism of a linear peristaltic pump is
shown in Figure 23-9. To create a linear peristaltic motion, cams with eccen-
tric axes are attached to a rotating cam shaft (Figure 23-10a) such that when
a shaft rotates, it moves the protruding finger up or down according to its
eccentric angle of rotation (Figure 23-lob).
 Biomedical Device Technology: Principles and Design

Flexible PVC tubing

(b)

___a Occlusion rollers

Direction of
fluid flow
protruding fingers
Figure 23-8. Peristaltic Infusion Mechanism. (a) Rotary Peristaltic; (b) Linear Peristaltic.

Fluid flow

Cams
Stepper motor and
reduction gear
Figure 23-9. Linear Peristaltic Pump Driving Mechanism.

Syringe Pumps
A syringe pump has a long screw mounted on the pump support. The
screw is rotated by a stepper motor and gear combination. The screw is sup-
ported by two bearings to allow smooth operation. As the screw rotates, it
moves a nut threaded onto the screw in the horizontal direction (Figure
23-11). The nut is attached to a pusher connecting to the plunger of a
 Infusion Devices

Protruding

,/ fingers
Rotating
I cams

-
/vw/ -
Cam rotation axis
(b)

Figure 23-10. Linear Peristaltic Plunger Driving Mechanism.

syringe, which is loaded with the solution to be infused. The flow rate of the
fluid coming out of the syringe depends on the rotational speed of the screw
and the cross sectional area of the syringe body. In mathematical terms, the
volume flow rate of a syringe pump is:

where:
F= volume flow rate in cubic centimeters per minute,
& rotational speed of the screw in revolutions per minute,
t = screw pitch in cm, and
A = cross-sectional area of the syringe plunger in cm'.
Syringe pumps are often used in high-accuracy, low-flow rate applica-
tions (e.g., 0.5 to 10 ml/hr) and when more uniform flow pattern is required.
It is also used to infuse thicker feeding solutions. A patient-controlled anal-
gesic (PCA) pump is a special syringe pump designed to allow patients to
self-administer boluses of narcotic analgesic for pain relief.
In general, piston cylinder infusion pumps and syringe pumps produce a
more accurate and consistent flow output. However, during low flow rate set-
tings, piston pumps (both piston cylinder and diaphragm) produce boluses of
infusion rather than a smooth flow pattern. Figure 23-12a shows the flow
pattern of a piston cylinder pump at a low flow rate setting (e.g., 10 ml/hr);
a bolus in the diagram corresponds to the flow of one stroke of the piston. A
bolus type of infusion may not be suitable for some applications such as
 Biomedical Device Technology: Princ$les and Design

Syringe with
Direction of infusion solution
plunger travel

Figure 23-1 1. Syringe Pump.

administering medications to small infants. Under ideal conditions, a syringe

pump will produce a uniform flow with little fluctuation. A peristaltic pump,
with its multiple protruding finger pumping action, produces a flow pattern
with less fluctuation than piston pumps as shown in Figure 23-12b.

n
- A
2- Boluses
Average flow rate
4
-3
(D

-.
5
7

10

Time (sec)

Average flow rate

Time (sec)

Figure 23-12. L o w Flow Rate Infusion Flow Pattern.

 Infision Devices

COMMON FEATURES

Some common features found in general-purpose infusion pumps are

described in the following sections. Depending on the design application, an
infusion pump may have other additional features.

Flow Rates
The flow rate of a general-purpose infusion pump can be set within a
range from 1 to 999 ml/hr with an accuracy of k5 to 10%. For neonatal
pumps, the range is 0.1 to 99 m/hr with an accuracy of 2%.+
Volume To Be Infused (VTBI)
A VTBI of 1 to 9,999 ml can be programmed such that the pump will
stop after this volume has been delivered. Usually, when VTBI is reached,
an audible tone will sound to alert the clinician. The pump will switch to its
KVO rate.

Keep Vein Open (KVO)

When infusion has stopped, in order to prevent blood clot at the
venipuncture site, a slow infusion rate of about 1 to 5 ml/hr is maintained to
flush the catheter to prevent blood clotting.

Occlusion Pressure Alarm

A pressure sensor inside the pump monitors the pressure of infusion. A
high pressure indicates occlusion downstream of the pump. An alarm is set
to notify the clinician to check the IV line. Downstream occlusion may be
due to clot IV catheter or pinching of the IV line (e.g., by the patient rolling
over the line). In some infusion pumps, the occlusion pressure alarm may be
adjusted to activate between 1 and 20 psi.

Fluid Depletion (or Upstream Occlusion) Alarm

When the IV bag is empty, a negative pressure will develop upstream of
the pump. An alarm to indicate such a condition can prevent air from enter-
ing the IV line and being infused into the patient.
 Biomedical Device Technology: Princ$les and Design

Infusion Runaway (or Free Flow) Prevention

Most modern pumps have a built-in mechanism to prevent free flow of
solution into the patient. Free flow can occur when the IV line is removed
from the pump while the occlusion clamp and roller clamp are both open.
When the pump is used to administer a potent drug to a patient, free flow
can impose serious risk to the patient if a large dose of such medication is
infused into the patient. A mechanical interlock on the IV line to shut off the
line when it is pulled out from the pump will prevent this.

Air-in-Line Detection
To prevent air embolism in patients, air-in-line detectors are built into
infusion pumps to detect air bubbles in the IV lines. Infusion will stop and
an alarm will sound when a large air bubble is detected during infusion.

Dose Error Reduction System

This is a software algorithm that checks programmed doses against pre-
set limits specific to certain drugs and clinical location profiles. It alerts clin-
icians if the programmed dose exceeded the preset limits. For example, drug
X used in area A has a dose limit of 20 mcg/kg/hr. If the dose, based on the
programmed flow rate and drug concentration, exceeds 20 mcg/kg/hr, infu-
sion will not start and an alarm will sound.

Battery Operation
Most pumps are powered by internal rechargeable batteries so that the
pump may be moved around with the patient during use. A battery low
detector circuit will alert the user if the battery is running low and must be
recharged.

FUNCTIONAL BLOCK DIAGRAM

Figure 23-13 shows a functional block diagram of a volumetric infusion

pump. The user input/output interfaces are shown on the left-hand side of
the diagram. User selectable inputs include:
Infusion flow rate
Volume to be infused (VTBI)
 Infision Devices

Keep vein open (KVO) enable selection

Pump start/stop control
The CPU, based on the input settings, controls the speed of the stepper
motor driving the pumping mechanism to deliver the set infusion rate. The
rotational speed of the pump driver is monitored by a LED/optical transis-
tor slit detector. Based on this rotational speed, the volume of infusion is
computed and compared to the VTBI setting. If KVO is enabled, the motor
speed will be reduced to the KVO rate when VTBI has been reached.
The pressure in the IV line is monitored by a pressure sensor pressing on
the IV tube inside the pump. When the pressure exceeds the occlusion pres-
sure, the CPU will shut down the pump and sound an alarm.
Air bubbles in the IV line are detected by an ultrasound transmitter and
receiver pair. The attenuation of ultrasound in air is higher than that in water.

Figure 23-13. Functional Block Diagram of Volumetric Infusion Pump.

 Biomedical Deuice Technology: Princ@les and Design

When an air bubble passes through the detector, the intensity of ultrasound
detected by the receiver will decrease. The duration of this decreased signal
corresponds to the size of the air bubble in the line. The CPU will stop infu-
sion and sound an alarm if a large air bubble is detected.

PERFORMANCE EVALUATION

An important performance parameter of an infusion device is its flow

rate accuracy. The flow rate of an infusion pump can be calculated by mea-
suring the volume of solution delivered over a period of time. For example,
a measuring cylinder can be used to collect the fluid infused over a period
of, say, 5 minutes at a certain flow rate setting. Such a method is generally
acceptable for most general-purpose infusion devices. However, this method
will not be appropriate to measure the accuracy of very low flow rate settings
due to the fact that it will take a very long time to collect enough solution to
obtain an accurate volume measurement (e.g., it takes 2 hours to collect 10
ml of solution at 5 ml/hr setting). Evaporation of fluid in the collection con-
tainer will also affect the accuracy of such low volume long duration mea-
surement. In addition, this measure gives only the average flow rate over a
period of time. No information regarding the flow pattern is obtained (flow
fluctuation, bolus effect, etc.).
Another parameter to be measured is the occlusion alarm pressure. This
pressure is measured by connecting the IV line to a pressure meter and then
starting the infusion. The pressure inside the line will quickly build up until
it reaches the occlusion pressure alarm limit. It is important to leave an air
buffer between the liquid line and the pressure meter should the meter not
be able to measure wet pressure.

Example
A measuring cylinder is used to collect fluid from an infusion pump during
a flow rate performance evaluation test. During the test, 9.6 mL of fluid is
collected over a period of 5 minutes. If the flow rate setting of the infusion
pump is 120 ml/hr, what is the accuracy of the pump?

Solution
From the test, 9.6 ml of fluid is infused in 5 minutes. Therefore, the calculat-
ed pump flow rate is 9.6 m1/5 min = 1.9 ml/min = 115 ml/hr. Therefore, the
percentage error of the infusion pump is 120 - X 100% = +4.2%
120
 Infision Devices

FACTORS AFFECTING FLOW ACCURACY

Other than electronic component failures and mechanical wear and tear,
the following common factors-affect the flow accuracy of infusion pumps.
Too high backpressure in the IV line can reduce the flow rate. Normal
backpressure depends on the flow rate, the diameter and length of the IV
tubing, and the viscosity of the IV fluid. The smaller the inside lumen and
the longer the tubing, the higher the backpressure. Backpressure increases
with increase flow and fluid viscosity. Backpressure may also be created
when the IV tube is kinked. When the backpressure is too high, the pump-
ing mechanism may not be able to overcome such pressure. For example,
during high backpressure, if the occlusion pressure created by the protrud-
ing fingers on the IV tubing is not high enough, fluid may leak backwards at
the location of occlusion.
Another potential problem associated with high backpressure is bolus
infusion. As the flexible IV tubing is slightly elastic, its diameter will increase
under high backpressure. Upon clearing the occlusion, the IV tubing will
recoil to its original diameter thereby releasing the stored fluid along the
length of the tubing. Therefore, a large bolus of fluid may be infused into the
patient.
For IV pumps using the peristaltic pumping mechanism, as the flow rate
depends on the inner diameter of the IV tubing, variation of the inner diam-
eter will change the rate of infusion. It is therefore important to ensure that
the inner diameter dimension of IV lines used with peristaltic infusion
pumps are manufactured within acceptable tolerance. In addition, as the sec-
tion of IV tubing under the protruding fingers is being compressed for a peri-
od of time with prolonged use, the shape and therefore the inner diameter of
the tubing will change. In order to avoid inaccuracy, manufacturers often
recommend that users move a different section of tubing under the pump-
ing mechanism every several hours.
Theoretically, a syringe pump should produce an accurate and uniform
flow pattern. In practice, however, under very low flow rate applications, the
plunger may stick to the side of the cylinder until the pusher delivers enough
force to overcome the static friction. Once the plunger is free, it will advance
rapidly and stop, thereby pushing a bolus of solution into the patient. This
sudden start and stop movement can repeat itself during low flow rate infu-
sion.
In general, among different pumping mechanisms of volumetric infusion
pumps, the piston cylinder pump is the most accurate but most expensive
due to the special infusion set with the piston cassette. The linear peristaltic
pump is very commonly used in general IV infusion since it has a fairly accu-
rate infusion rate. In addition, most peristaltic pumps can use ordinary grav-
 Biomedical Device techno lo^: Pfinc@les and Design
ity infusion sets and are therefore less expensive to operate than those that
require dedicated infusion sets.
 Chapter 24

ELECTROSURGICAL UNITS

OBJECTIVES

Describe the tissue response to electrosurgical current in terms of des-

iccation, fulguration, and cutting.
Identify the characteristics of the cut, coagulation, and blended electro-
surgical waveforms.
Explain the constructions and functions of active bipolar and monopo-
lar electrodes, and the dispersive return electrode.
Sketch the block diagram of an electrosurgical generator and explain
the functions of each block.
Analyze ESU waveform generation circuits.
Perform quality assurance testing on an ESU generator.
Identify the potential hazards of electrosurgery and safety precautions
during electrosurgical procedures.

CHAPTER CONTENTS

1. Introduction
2. Principle of Operation
3. Modes of Electrosurgery
4. Active Electrodes
5. Return Electrodes
6. Functional Building Blocks and ESU Generators
7. Output Characteristics
8. Quality Assurance
9. Common Problems
 Biomedical Device Technology: Princ$les and Design

An electrosurgical unit (ESU) delivers high-frequency electrical current

through an active electrode to produce cutting and coagulation effects on tis-
sues. The frequency of electrosurgery for cutting is between 100 kHz and 5
MHz, which is within the radio frequency (RF) band. When appropriately
modulated, this RF frequency can cut through tissues and cauterize bleeding
blood vessels. The simultaneous cutting and hemostatic effect make ESU
useful for procedures on tissues with capillaries and on patients receiving a
high dose of anticoagulant drugs.
Electrosurgical units using spark-gap generators have been used since the
1920s. Most ESUs today use solid-state technology with microprocessor-con-
trolled output for better results and improved safety. Argon-enhanced ESU
systems, using a jet of argon gas to cover the surgical site during electro-
surgery, can provide rapid and uniform coagulation over a large bleeding
surface and are therefore useful in surgeries on organs such as the liver,
spleen, and lung. ESUs are also used in endoscopic procedures to perform
ablation, desiccation, cauterization, and removal of tissues. Special hand-
pieces are designed for different ESU procedures.

PRINCIPLE OF OPERATION

An ESU is a RF generator. Two electrodes, one called active and the

other called passive, are used to apply the RF current from the ESU to the
patient. At the beginning of the procedure, the passive electrode (or return
electrode) is attached to the patient's body. The surgeon then applies the
active electrode to the surgical site to achieve the surgical effect. The active
electrode is usually a very small tip electrode, while the return electrode has
a large contact surface area with the patient. The high-frequency current
passing through the tissue creates the surgical effect. The surgical effect is due
to heat created by the RF current at the tissue-active electrode interface. The
degree of heating in the tissue depends on the resistivity of the tissue as well
as on the RF current density (the resistivity of soft tissue is about 200 a m ) .
Figure 24-1 shows a typical setup of an electrosurgical procedure.
Different current densities create different effects on living tissues. Table
24-1 shows typical tissue effects at different levels of RF current density.
In practice, a current density much higher than 400 mA/cm2 at the sur-
gical site is necessary to produce electrosurgical effect. However, to prevent
tissue injury beyond the surgical site, the current density must be limited to
below 50 rnA/cm2 in tissues outside the surgical site, This is a&ieved hy
 Electrosurgical Units

Active electrode

Dispersive current

Return electrode
Figure 24-1. Electrosurgery Setup.

Table 24-1.
-- -
Tissue Effect of RF Current Density.
-

Current Density Tissue Effect

> 50 mA/cm2 Reddening of tissue
> 80 mA/cmi Pain and blistering
Intense pain
second-degree burn

attaching a large surface area electrode on the opposite side of the active
electrode so that the return current is dispersed over a larger area within the
patient's body. Figure 24-1 shows the active electrode applied to the surface
of the tissue and the flow of current inside the tissue when the return elec-
trode is placed far from the active electrode. The density of the RF current
flowing in the tissue closest to the active electrode is the highest, and it
decreases rapidly (inversely proportional to the square of the distance from
the surgical site) at locations farther from the active electrode site.
Three different tissue effects can be created by an electrosurgical current
at the active electrode site: desiccation, cut, and fulguration.

Desiccation
When a relatively small RF current flows through the tissue, it produces
heat and raises the tissue temperature at the surgical site. Heat will destroy
and dry out the cells. This process may produce steam and bubbles and
eventually turns the tissue a brownish color. This mechanism of tissue dam-
age is called desiccation. It is achieved by placing the active electrode in con-
 Biomedical Device Technology: Principles and Design

tact with the tissue and setting the ESU output to low power. As desiccation
is created by the heating (IZR)effect, any current waveform may be used for
desiccation.

Cut
By separating the active electrode by a small distance (about 1 mm) from
the tissue and maintaining a few hundred volts or higher between the active
and return electrodes, RF current may jump across the separation, produc-
ing sparks. Sparking creates intense heat, causing cells to explode. Such
destruction of cells leaves behind a cavity. When the active electrode moves
across the tissue, this continuous sparking creates an incision on the tissue to
achieve the cutting effect. In general, a high-frequency (e.g., 500 kHz) con-
tinuous sine wave is used to create the cutting effect. Cutting usually requires
a high power output setting.

Fulguration
To produce fulguration, the energized active electrode first touches the
tissue and then withdraws a few millimeters to create an air gap separation.
As the active electrode moves away from the tissue, the high voltage creates
an electric arc jumping across the active electrode and the tissue. This long
arc burns and drives the current deep into the tissue. Intermittent sparking
does not produce enough heat to explode cells, but it causes cell necrosis and
tissue charring at the surgical site. Fulguration coagulates blood and seals
lymphatic vessels. To achieve fulguration, most manufacturers use bursts of
a short-duration damped sinusoidal waveform. The sinusoidal waveform is
usually the same frequency used for cutting (e.g., 500 kHz), and the repeti-
tion frequency for the bursts is much lower (e.g., 30 kHz). A higher voltage
waveform is required to maintain the long sparks. Although the peak voltage
is higher, fulguration produces less power than cutting due to its low duty
cycle.
Table 24-2 summarizes the three mechanisms of electrosurgery.

MODES OF ELECl'ROSURGERY

The cut mode in electrosurgery applies a continuous RF waveform (sinu-

soidal or near sinusoidal) between the active and return electrodes. The
coagulation mode uses bursts of a higher voltage damped RF sinusoidal
waveform (to create fulguration tissue effect). Instead of switching back and
 Electrosurgical Units

Table 24-2.
Mechanism of Electrosurgery. --
Tissue Effect --
Active Electrode Power
Desiccation Heat dries up tissue, produces Monopolar or bipolar. Low
steam and bubbles. In contact with tissue.
Turns tissue brown.
Cut Sparking produces intense heat, Monopolar. High
explodes cells leaving cavity. Electrode separated from
Incision on tissue caused by tissue by a thin layer of
continuous sparking. steam.
Fulguration Intermittent sparking does not Monopolar. Medium
produce enough heat to explode Electrode separated by
cells. Heat causes necrosis to an air gap.
tissue.
High voltage drives current deep
into tissue, chars tissue to carbon
- -

forth between cut and coagulation during a procedure, most ESUs have one
or more blended modes, which allow simultaneous cutting and coagulation.
A blended waveform has a lower voltage level but a larger duty cycle than
the coagulation waveform. Figure 24-2 shows an example of the cut, blend-
ed, and coagulation output waveforms of an ESU. A blended mode with a
larger duty cycle will have more cutting effect than one with a lower duty
cycle.
The setup shown in Figure 24-1 with the active electrode and the large
surface return electrode is called a monopolar operation. Instead of placing
a separate return electrode away from the surgical site, a bipolar handpiece
has both the active and return electrodes grouped together (e-g., an ESU for-
ceps). Biopolar electrodes are often used to perform localized desiccation on
tissue. In Figure 24-3, the ESU is switched to bipolar coagulation mode to
cauterize a section of a blood vessel before it is cut apart to avoid profuse
bleeding.
Table 24-3 lists the characteristics of different modes of ESU operations.
The crest factor (last column) is defined as the peak voltage amplitude of the
ESU waveform divided by its root mean square voltage. For a continuous
sine wave, the crest factor is 1.41. Since a pure sine wave has little or no
hemostatic effect on tissues, most manufacturers use a lightly modulated sine
wave to achieve a small degree of hemostatic effect in the cut mode. The
crest factor of the coagulation waveform is the highest (about 9) since it has
the largest peak voltage but the smallest duty cycle. In general, the higher the
crest factor, the more hemostatic effect the ESU waveform will have on tis-
sues.
 Biomedical Device Technology: Principles and Design

Figure 24-2. ESU Output Waveforms. (a) Cut, (b) Blended, (c) Coagulation.

Segment of blood
vessel to be coagulated
Figure 24-3. Bipolar Mode of Electrosurgery.
 Electrosurgical Units

Table 24-3.
--
Characteristics of
- .
-
ESU Operation
-.
Modes.
- . -

-
Efect -.
Waveform
-
Eltuge Power Crest Factor
Monopolar
Cut Pure incision Continuous Low High - 1.41 to 2
plus slight unmodulated
hemostatic effect sine wave to
lightly modulated
sine wave
Coagulation Desiccation or Burst of damped High Low
fulguration sine wave
Blended Cut and Burst of medium Medium Medium Between
coagulation duty factor cut and
sine wave coagulation
Bipolar
Coagulation Desiccation Continuous Lowest Lowest 1.41
unmodulated
sine wave
-

ACTIVE ELECI'RODES

ESU active electrodes for monopolar operations come in different forms

and shapes. The most common active electrode is the flat blade electrode,
which can be used to perform cutting and coagulation. Some of the other
commonly used active electrodes are needle, ball, and loop electrodes. Ball
electrodes are usually used for desiccation (by pressing the electrode against
the tissue and passing the RF current through the tissue). Loop electrode,
with its conductive wire loop, is used to remove protruded tissues such as a
nodule. The metal tips of the electrodes (Figure 24-4b) are usually single-use
disposable units. The electrode handles may be multiple use or single use.
The handle part of the electrode may have a switch to activate the ESU. A
foot switch operated by the surgeon may be used instead of the hand switch.
The combination of an ESU handle and a tip (Figure 24-4a) is often referred
to as an ESU pencil (or a hand-switched ESU pencil if a switch is located on
the handle).

RETURN ELECI'RODES

While the function of the active electrode is to create the surgical effects,
the return electrode (or passive electrode) in monpolar ESU operation pro-
 Biomedical Device Technology: Princ$les and Design

Hand activation switch

Active electrode
/

(a)
Cable to ESU

Figure 24-4. (a) Hand-Switched ESU Pencil with a Flat Blade Electrode;
(b) Monopolar Tips: 1) Loop, 2) Flat Blade, 3) Needle, 4) Ball.

vides the return path for the ESU current. As mentioned earlier, the maxi-
mum RF current density level to avoid causing any tissue damage is 50
mA/cm2. A large surface area electrode (e.g., 100 cmL)is therefore required
to limit the current density below this safe level in tissues away from the sur-
gical site, including those in contact with the return electrode.
There are many types of return electrodes for ESU procedures. Bare
metal plates placed under and in contact with the patient were used in early
days. However, it was noted that burns (primarily heat burns) and tissue
damage sometimes occurred at the return electrode sites. Investigations
revealed that the primary cause of such patient injuries was due to poor elec-
trode-skin contact or insufficient contact surface area between the electrode
and the patient a artof the electrode not in contact with the ~atient).It was
also noted that burns often appeared in the form of rings at the skin surface.
Laboratory experiments showed that the current density at the skin-return
electrode interface is highest around the rim of the electrode. Figure 24-5
shows the current density distribution of such an experiment. This occur-
rence is due to the fact that electrons are negatively charged particles; when
they are allowed to freely move in a conductive medium, they will repel each
other and therefore more will end up at the perimeter of the medium, in this
case at the perimeter of the return electrode. This phenomenon is known as
the "skin effect" in electrical engineering, where the current density of high-
frequency current in a conductor is very much higher at the surface of the
conductor than in its core.
Today, conductive gel pads are used for ESU return electrodes. A con-
ductive gel pad electrode has a self-adhesive surface to avoid shift and falloff
and is flexible to fit the contour of the patient's body. Return electrodes are
 Electrosurgical Units

designed so that, under normal use, no skin burn will occur at the return
electrode site. To ensure patient safety, technical standards are in place spec-
ifying the performance of return electrodes. For example, the ANSI/AAMI
HF18 Standards stipulate that the overall tissue-return electrode contact
resistance shall be below 75 R. In addition, no part of the tissue in contact
with the return electrode shall have more than a 6OC temperature increase
when the ESU is activated continuously for up to 60 seconds with output cur-
rent up to 700 mA.

I
...-. . ....,, I Return electrode
.......- ...........................................-.............
...... .....................................
......
I I
I
2 A
7
7
I
I
I
[
I
Skin surface
(D
2
a
(D
3
cn
2

I I
b
I I
I I Distance
Figure 24-5. Current Density Crossing the Return Electrode-Skin Interface.

Due to problems associated with bums, special monitoring devices are

often built into ESUs to monitor the integrity of the return electrode path. If
the integrity is breached, an alarm will sound and the ESU output will be dis-
abled to prevent patient injury. Two levels of monitoring are often available
for high output power ESU (e.g., output greater than 50 W). The first is
return electrode monitoring and the second is return electrode quality mon-
itoring.

Return Electrode Monitor (REM)

A REM system monitors the return path of the electrode to the ESU. It
detects the continuity of the return electrode cable from the electrode to the
ESU. In a typical REM system, a double conductor cable and a low-fre-
quency (relative to the ESU output frequency; e.g., 140 KHz) low-current
(e.g., 3 mA) isolated source from the ESU are used to measure the resistance
 Biomedical Device Technology: Principles and Desip

of the return cables (Figure 24-6a). A high resistance (e.g., > 20 R) will trig-
ger the REM alarm.

Return Electrode Quality Monitor (REQM)

REM measures only the continuity of the return electrode cable, not the
quality of contact between the electrode and the patient, whereas REQM
monitors both. Figure 24-6b illustrates the principle of the REQM. In
REQM, a dual conductive pad electrode is used. The right-hand side dia-
gram in Figure 24-613 shows the cross-sectional view of the electrode-skin
interface. The small monitoring current flows from the ESU REQM circuit
to one of the conductive pads, passes through the two electrode-skin inter-
faces, and returns to the ESU via the second pad. Too high a REQM resis-
tance (e.g., greater than 135 R) suggests poor electrode-skin contact or open
circuit return cable; too low a resistance (e.g., less than 5 R) suggests a short
circuit between the two conductive pads. In addition, some machines may
sound an alarm if the REQM detects a large change in the resistance (e.g.,

Cable to ESU
REM connector

Adhesive gel pad

REQM connector
Figure 24-6. Return Electrode and Return Electrode Quality Monitors.
 Electrosurgical Units

resistance increase by more than 40% from the initial reference value).
FUNCTIONAL BUILDING BLOCKS OF ESU GENERATORS

The spark gap ESU generator developed in the 1920s consists of a step-
up transformer T1, which increases the 60 Hz 120 V line voltage to about
2,000 to 3,000 V (Figure 24-7). As the sinusoidal voltage at the secondary of
T1 increases from zero, electrical charge accumulates in the capacitor C1
and the gas inside the spark gap (a gas discharge tube) starts to ionize until
an arc is formed between its electrodes. Arcing (or sparking) of the spark gap
resembles closing of a switch in the series resonance circuit formed by C1,
L1, and the impedance of the spark gap. The fundamental frequency of the
arcing current is approximately equal to the resonance frequency of Ll/Cl.
The voltage amplitude of this high-frequency oscillation will decay until the
arc is extinguished. Proper choice of L1 and C1 produces an RF damped
sinusoidal waveform that occurs twice within one period of the 60 Hz input
signal. This RF damped sinusoidal waveform is coupled to the output circuit
by induction between Ll and L2. The output level is selected by the taps
selection on L2. The RF chokes L3 and L4 (or RF shunt capacitor C4) are
used to block the RF signal from entering the power supply. Spark gap gen-

- Activation switch Output power selector

C1

120 V 60 Hz
AC power L1 Output
C4
Spark gap
L4

T1
-
-
-L

Figure 24-7. Spark Gap ESU Generator.

erators are primarily used for coagulation or cauterization.

Spark gap ESUs were commonly used until the early 1980s, when they
began to be replaced by solid-state generators. In a solid-state ESU, the RF
frequency (e.g., 500 kHz) and the burst repetition frequency (e.g., 30 kHz)
are generated by solid-state oscillators. The shape of the ESU waveform (cut,
blended, or coagulation) is created by combining the frequencies of these
two oscillators. The waveform is then amplified by a power amplifier and a
 Biomedical Device Echnolo~:Princ$les and Design

step-up transformer. The output of an ESU can go up to 1,000 watts, 9,000

volts (peak to peak open circuit voltage), and 10 amperes. Figure 24-8 shows

rnoscillator

-
Coagulation

+ +
oscillator

Gating
and wave
shaping
+
Output
level
control
+
Power
ampllfler
+
Output
transformer
circuit

To ESU active
and return
electrodes
v
User Input
Audlo and
vlsual
lndlcators

Figure 24-8. Functional Block Diagram of an ESU.

the simplified functional block diagram of a solid-state ESU.

The output stage of an ESU is shown in Figure 24-9. In the circuit, the
waveform created by the gating and wave shaping circuit is fed into the base
of a power amplifier Ql, and the output of the amplifier connects to the pri-
mary winding of the output transformer TI. The transformer and the power
amplifier circuit are connected to a 200 V DC power source. For high-power
output ESUs, a number of power transistors connected in parallel form the
output circuit. Each of these transistors shares a portion of the output power.
The ESU waveform, after steps up by the output transformer to several thou-
sands volts, is fed across the active and return electrodes via a pair of capac-
itors C1 and C2. These capacitors behave like a short circuit to RF but block
low-frequency (60 Hz) leakage current to the patient.
The ESU output circuit shown in Figure 24-9 is considered an isolated
output ESU as there is no connection from the patient circuit (the secondary
of the output transformer) to the power ground. Theoretically, for an isolat-
ed output ESU, a person touching the active electrode but not the return
electrode will not get a shock or bum when the ESU is energized. However,
due to the high'frequency and nonzero leakage capacitance, if the person
also touches a grounded object, some F W current will flow from the active
 Electrosurgical Units

active electrode

return electrode
ESU waveform from T1
gat~ngand wave C2
shapingcircuit
Q1

-
-
Figure 24-9. ESU Output Circuit.

electrode to the person and return to the ESU via this ground leakage path.
High-frequency leakage current may be on the order of magnitude of a few
tens of rnA.

OUTPUT CHARACTERISTICS

Table 24-4 lists the output characteristics of a typical ESU. Figure 24-10
illustrates the output characteristics of the ESU cut waveform at different val-
ues of patient load. Note that according to the output characteristics, the ESU
is rated to produce 300 W of output only when the patient load is at 300 a.
The output power is reduced to 180 W when the patient load becomes 800
IR. According to the ESU output characteristics, the output power decreases
as the patient load increases. As the tissue impedance depends on the type
of tissue as well as the condition of the tissue, this may create problems dur-
ing the operation as the output power at a particular setting will fluctuate
with the tissue impedance. To overcome this problem, some manufacturers
have produced ESUs that can measure the tissue impedance and automati-
cally restore the output power to the set value.
In most electrosurgical procedures, the active electrode is energized only
 Biomedical Device Technology: Principles and Design

Table 24-4.
- --- -.
ESU Output Characteristics.
pp .
- - - - - - - -- - -
Mode Waveform Max. P-P Rated Patient Output Power
Open Circuit Load (a) (at rated load)
- - - - - - - - - -
Voltage (V3
- - - - - - --
(w)
-

Cut 500 kHz sinusoidal 3,000 300 300

Blend 1 500 kHz burst of sinusoidal at
5O0/)duty cycle repeating at
30 kHz
Blend 2 500 kHz burst of sinusoidal at
37.5(%/0
duty cycle repeating at
31 kHz
Blend 3 500 kHz burst of sinusoidal at
25(%1duty cycle repeating at
30 kHz
Coagulation 500 kHz burst of damped 7,000 300 120
sinusoidal repeating at 30 kHz
Bipolar 500 kHz sinusoidal 800 100 70
- -

Figure 24-10. ESU Cut Mode Output Characteristics.

intermittently and each activation lasts for a short period of time (e.g., 15 sec-
onds for cutting in general surgery). Table 24-4 shows the peak to peak open
circuit voltage of different modes of operation. However, when the current
starts to flow (i.e., an arc has been established), the voltage across the active
and return electrodes will drop substantially.
 Electrosurgical Units

QUALITY ASSURANCE

Since an ESU delivers high-energy therapeutic current, it is essential to

ensure that the machine is safe and is operating according to its designed
specifications. Other than general electrical safety inspection, the following
performance tests should be carried out periodically.

Output Power Verification Test

The output power of an ESU should be measured against the manufac-
turer's specifications. Figure 24-1 1 shows the setup to measure the ESU out-
put power. The output waveform can be sampled across the sample resistor
Rs and displayed on the oscilloscope. The output voltage Vo of the ESU is
calculated from the resistance values by the equation:

vo = R Rs
+Rs Vs.

If the output voltage is a sine wave, the power output may be calculated from
the equation:

P= vo'
R+Rs '

+
Note that the load resistance RL is equal to (R Rs) and both should be spe-
cial noninductive resistors and of sufficient power rating to withstand the

ESU

Figure 24-11. ESU Output Power Measurement.

406 Biomedical Device Technology: Principles and Design

ESU output.
High-Frequency Leakage Test
High-frequency leakage refers to the current flowing from either the
active electrode to ground or the return electrode to ground when the ESU
output is activated. Ideally, the amount of leakage current should be zero.
However, due to the nature of the high frequency, a significant amount of
capacitive leakage current will flow between the active electrode and ground
as well as between the return electrode and ground. Figure 24-12b shows the
setup to measure the high-frequency leakage from the active electrode to
ground. To measure the leakage from the return electrode to ground, the
load resistor is connected to the return electrode connection of the ESU and
the active electrode connection is left open. The allowable leakage found by
measuring the power dissipated by the load resistance RL (e.g., less than 4.5
W for RL = 200 a). Percentage isolation is a common value to represent the
degree of isolation. It is defined as:

OIOIsolation = (1 - P isolation ) X 100%.

P normal

ESU I ESU

RL
#tVL
Pisolation

Figure 24-12. ESU Isolation Test.

Some manufacturers (and standards) call for the %Isolation to be greater

that 80% for a load resistance RL within the range of 100 to 1,000 a.
Special testers with built-in potential dividers, variable patient load, and
switchable configurations are available to facilitate these measurements.

COMMON PROBLEMS
 Electrosurgical Unih

Problems associated with electrosurgery may be grouped into four dif-

ferent categories:
Burns
Fire
Musclehewe stimulation
Electromagnetic interference

Burns
Skin burns at the return electrode site are one of the more common safe-
ty problems for patients under electrosurgical procedures. The main causes
of skin burn are poor electrode-skin contact, inadequate site preparation,
and pressure points on the electrode contact surface (which creates a low
resistance pathway for the current).
Internal tissue burns are caused by the concentration of ESU current
along a low resistance path such as a metal implant or a pacemaker
lead wire near the active or return electrodes sites.
For grounded ESUs or ESUs with isolation failure, RF current may
flow through a secondary ground path on the patient (e.g., a patient's
arm may receive a burn at the location where it is touching a ground-
ed object).
In endoscopic or laparoscopic procedures, an insulation failure on the
shaft of the ESU handpiece will cause tissue bum when such failure
creates a secondary conduction path between the active electrode and
the tissue.
Too high a power setting and too long an activation period (e.g., dur-
ing a liver tumor ablation procedure) when an undersized return elec-
trode was used or the return electrode was not properly applied.
Patient or staff bums by an activated ESU pencil when it was inad-
vertently energized (e.g., someone accidentally stepped on the ESU
foot activation switch) while touching the patient or a staff member.

Fire and Explosion

An electrosurgical procedure produces sparks and arcing. The sparks
may ignite flammable materials such as body hair, cotton drapes, or a pool
of alcohol used for disinfection. This situation is worsened under an enriched
oxygen environment, which is commonly found in operating room areas.
There have been incident reports on cases of explosion inside the
abdominal cavity when the ESU ignited flammable bowel gas inside the
patient.
 Biomedical Device Technology:Princ$les and Design

Muscle and Nerve Stimulation

The reason ESU frequency is above 100 kHz is to avoid muscle and
nerve stimulation. Under normal circumstances, muscle and nerve fibers are
not triggered by current lower than 100 kHz. However, studies have shown
that arcing may produce lower frequency current, which can stimulate nerve
or muscle fibers. As a precaution, it is contraindicated to perform electro-
surgery near major nerve fibers.

Electromagnetic Interference (EMI)

An ESU is a RF source. Although the machine may be shielded to pre-
vent radiation and conduction of EMI, the electrodes and cables act as
antennae to broadcast the RF frequencies. Older medical devices, with lower
electromagnetic immunity, can be adversely affected by the EM1 from elec-
trosurgery. Devices may reset, produce errors, or switch to another mode of
operation under EM1 influence. Improperly grounded devices are especial-
ly vulnerable to EMI.
 Chapter 25

RESPIRATION MONITORS

OBJECTIVES

Explain the mechanics of breathing.

Describe common respiration parameters.
Explain the principle of operation and construction of medical spirom-
eters.
Define standardized gas volume in BTPS.
Explain the principles of respiration monitoring using the impedance
pneumographic and thermistor methods.
Sketch a block diagram of a respiration monitor using the method of
impedance pneumography and explain the functions of each block.
List factors affecting signal quality, accuracy, and patient safety in res-
piratory monitoring.

CHAPTER CONTENTS

1. Introduction
2. Mechanics of Breathing
3. Parameters of Respiration
4. Spirometers
5. Respiration Monitors
 Biomedical Device Technology: Princ$les and Design

INTRODUCTION

The primary function of the lungs is to exchange gas between the

inspired air and the venous blood. Air is inhaled by voluntary or involuntary
action and is presented to one side of the membrane of the alveoli; venous
blood is on the other side of the membrane. Gas exchange between air and
blood occurs across this membrane. In the process, carbon dioxide is
removed and oxygen is introduced into the bloodstream. For a normal adult,
this gas-blood barrier is less than 1 pm and has a total surface area of about
100 m2.
Disturbances of the respiratory system can be caused by a number of fac-
tors within the system or by other disorders. Diagnosis of respiratory disor-
ders therefore can provide information about the well-being of the respira-
tory system as well as other organs or body functions. The rhythmic action
of breathing is initiated in the respiration centers of the pons and medulla.
The level and rate of respiration are controlled by the partial pressure of car-
bon dioxide and oxygen as well as the pH of the arterial blood. For exam-
ple, a decrease in blood pH (e.g., due to increases in metabolism), increased
accumulation of carbon dioxide in the arterial blood, and arterial hypoxemia
will increase ventilation.
This chapter introduces some methods of monitoring respiration pattern
and parameters in the clinical environment.

MECHANICS OF BREATHING

The lung is elastic and will collapse if it is not held expanded. At the end
of expiration or inspiration, the pressure inside the lung (or alveolar pres-
sure) is the same as the atmospheric pressure, whereas the pressure outside
the lung (or intrapleural pressure) is below atmospheric pressure. This nega-
tive pressure keeps the lung inflated. If air is introduced into the intrapleur-
a1 space, the lung will collapse and the chest wall will move outward. This
disorder is called pneumothorax.
The most important muscle for inspiration is the diaphragm. When it
contracts, the abdominal contents are forced downward and forward. This
action increases the vertical dimension of the chest cavity. In addition, the
external intercostal muscles contract and pull the ribs upward and forward,
causing a widening of the transverse diameter of the thorax. In normal tidal
breath (or passive breathing), the diaphragm descends by about 1 cm, but in
forced breathing, a total descent of up to 10 cm may occur. Under active
breathing (e.g., during heavy exercise), the abdominal muscles play an
 Respiration Monitors

important role in expiration by pushing the diaphragm upward. The internal

intercostal muscles assist active expiration by pulling the ribs downward and
inward, thus further decreasing the volume of the thoracic cavity. Diseases
that cause problems in these muscles or the nerves that innervate these mus-
cles create disorders in breathing.

PARAMETERS OF RESPIRATION

Parameters of respiration include lung capacities, respiration rate,

intrathoracic pressure, airway resistance, and lung compliance (or lung elas-
ticity). In addition to these parameters, respiration waveform as well as end-
tidal carbon dioxide concentration and its variations can all provide useful
information in the assessment and disease diagnosis of the respiratory and
related systems.
Figure 25-1 shows the volumes and capacities of respiration. The tidal
volume (TV) measures the volume of inspired or expired gas during normal
breathing. It is about 500 ml for a normal adult at rest. Inspiratory reserve
volume (IRV) is the maximum amount of gas that can be inspired from the
end-inspiratory level (or peak of the tidal volume). The sum of TV and IRV
forms the inspiratory capacity (IC).The expiratory reserve volume (ERV) is
the maximum amount of gas that can be exhaled from the end-expiratory
level (or trough of the tidal volume). The sum of IC and ERV, which is the
maximum volume of gas that the lung can expel or inhale, is the vital capac-
ity (VC). The residual volume (RV) is the amount of gas remaining in the
lung at the end of maximum expiration. This is the amount of gas that can-
not be squeezed out, or anatomic dead space, of the lung. The total lung
capacity (TLC) is the sum of RV and VC. These parameters, as well as the
ratios of some of them (e.g., RV/TLC), are used to assess the healthiness of
the respiratory system.
Airway resistance measures the ease of airflow during inspiration and
expiration through the bronchi and bronchioles. This is expressed as the
pressure difference between the mouth and the alveoli per unit of air flow.
The normal value is about 1 to 2 cmHzO per liter per second of flow at nor-
mal flow rate. This resistance becomes higher at higher flow rates. Lung
compliance measures the ability of the alveoli to expand and recoil to its
original state during inspiration and expiration. The normal lung at rest
expands by about 200 ml when the intrapleural pressure falls by 1 cmHnO.
Therefore, the compliance of the lung is 200 ml/cmHnO. At high lung vol-
umes, the lung is less easy to expand and thus its compliance falls. When the
airway is restricted, the air resistance increases. When the lung becomes
more fibrous, it loses its compliance.
 Biomedical Deuice Technology: Princ$les and Design

To produce work to move the chest wall and force air along the airways,
the respiratory muscle must consume oxygen. In normal subjects, the work
of breathing is very small except in large ventilation during heavy exercise.
However, in patients with obstructive lung disease, the resistance to airflow
is very high even at rest and therefore the work of breathing can be 5 or 10
times its normal value. Under these conditions, the oxygen cost of breathing
may become a significant fraction of the total oxygen consumption. Patients
with a reduced compliance of the lung also have a higher work of breathing
due to the stiffer structures. These patients tend to use shallower but more
frequent breaths to reduce their oxygen cost of ventilation. However, the air
exchange is not efficient in shallow breathing due to the fixed volume of air
in the anatomic dead space in the lung, bronchi, and bronchioles.
One of the most useful tests in a pulmonary function laboratory is the
analysis of a single forced expiration. The patient makes a full inspiration
and then exhales as hard and as fast as possible into a spirometer (a flow and
volume measurement device). The volume measured is called the forced
vital capacity (FVC). FVC is usually less than the vital capacity (VC), which
is obtained at slow expiration. The volume exhaled within the first 1 second
is called the forced expiratory volume, or FEVI. In obstructive lung disease
 Respiration Monitors

(such as emphysema), due to high airway resistance, both FEVl and the ratio
FEVdFVC are reduced. In restrictive lung disease (such as sarcoidosis), due
to the limited lung expansion, FVC is low, but because the airway resistance
is normal, the ratio FEVdFVC is high. Another index that can be derived
from a forced expiration is the maximal midexpiratory flow (FEF25-7.51~,),
which is obtained by dividing the volume between 75% and 25% of the FVC
by the corresponding elapsed time. This is a sensitive parameter to detect air-
way obstruction in early chronic obstructive lung disease. Figure 25-2 shows
typical records of forced expiratory volume measurements. FEVi is a useful
screening procedure to assess lung function and the efficacy of bronchodila-
tor therapy, and in following the progress of patients with asthma or chron-

- -
ic obstructive lung disease.

Normal Obstructive Restrictive

1 sec 1 sec 1 sec
++ Time

Figure 25-2. Forced Expiratory Volume Measurement.

The functional residual volume (FRV)is the volume of air in the lungs at
the end of expiration which is also the volume of air remaining in the lungs
between breaths. It is an important lung function as it changes markedly in
some pulmonary diseases. FRV is measured using an indirect method called
the helium dilution method. In this method, a container of known volume
(q is filled with a mixture of air and helium of concentration Ci~e.The
patient first breathes normally for a few cycles. At the end of the last expira-
tion (the volume of gas inside the lungs is the FRV), the patient starts and
continues to breathe from the container. After several breaths, the gas in the
container is diluted and mixed thoroughly with the gas in the lung. If the
helium concentration of the mixed gas is Cme, FRV can be calculated from
the equation:
 Biomedical Device Technology: Princ$les and Design

During inspiration, some of the air that a person breathes never reaches
the alveoli for gas exchange to take place. This air is called the dead space
air. During expiration, this volume of air expires to the atmosphere before
the air from the alveoli. The nitrogen method is used to measure the dead
space volume. In this method, the patient first breathes normal air and sud-
denly takes a breath of pure oxygen. This intake of pure oxygen fills the
entire dead space volume and some mixes with the alveolar air. The patient
then expires through a nitrogen meter to produce a nitrogen concentration
curve as shown in Figure 25-3. The initial expired air that comes from the
dead space consists of pure oxygen. After a while, when the alveolar air
reaches the nitrogen meter, the nitrogen concentration rises and then levels
off. The concentration of nitrogen is plotted against the volume of expired
air. The measurement terminates at the end of the expiration. The nitrogen
concentration curve divides the graph into two regions with areas A1 and A2
as shown. One can derive that the volume of dead space air VDcan be cal-
culated from the equation:

where VE= the total volume of the expired air.

Volume of
expired air (ml)

Figure 25-3. Nitrogen Dead Space Volume Measurement.

 Respiration Monitors

SPIROMETERS

A spirometer is a device to measure the flow and volume of gas moving

in and out of the lungs during inspiration and expiration. There are two cat-
egories of spirometers: one senses volume and the other senses flow. A vol-
ume-sensing spirometer has a container to measure the gas volumes; gas flow
rate can be calculated from the volume-time information. A flow-sensing
spirometer has a flow transducer placed in the gas flow pathway to measure
the flow rate of gas. Gas volume can be derived from the flow-time informa-
tion. Figure 25-4 shows the functional block diagram of a spirometer. The
patient circuit allows the patient to breathe in and out of the spirometer; the
transducer converts the volume or flow to an electrical signal. The processor
computes the respiratory parameters from the collected information and dis-
plays them on the output device such as a visual display or paper chart
recorder.

Expiration Patient Flow or

Inspiration + circuit + volume + Processor + Display
transducer

Figure 25-4. Block Diagram of a Spirometer.

Volume Transducers
Three commonly used volume transducers for respiration measurement
are shown in Figures 25-5 and 25-6. The water-sealed inverted bell spirom-
eter (Figure 25-5) moves up and down according to the respiration of the
patient. The low friction water seal and the counterweight attached to the
inverted bell reduce the resistance and backpressure, thereby allowing accu-
rate volume and flow measurements. A pen, which writes on a rotating
drum, is mechanically linked to the inverted bell. A rolling seal with a hori-
zontally mounted bell (Figure 25-6a) can also be used as the volume trans-
ducer in a spirometer. The horizontal mounting of the bell eliminates the
need for a counterweight and therefore simplifies the construction of the
spirometer. A third type of volume-measuring transducer is a bellow (Figure
25-6b). As gas moves in and out of the bellow, it inflates or deflates the bel-
low. The moving bellow can move a pen to record the changing volume on
a paper chart.
 Biomedical Device Technology: Principles and Design

Inverted bell
Rotating paper

Figure 25-5. Water-Sealed Inverted Bell Spirometer.

To patlent
mouthpiece
To patient
mouthp~ece

,
+
(a) Rolling Seal Volume Transducer (b) Bellow Volume Transducer

Figure 25-6. Volume Transducers.

Flow Transducers
Spirometers using flow transducers with no moving parts are commonly
used to minimize errors due to mechanical wear and tear. Figure 25-7 shows
the block diagram of such a spirometer. Many different flow transducers can
be used; examples are hot air anemometer, differential pressure flow trans-
ducer, et cetera. (see Chapter 7 for principles of flow transducers). Modern
spirometers are microprocessor-based and have built-in compensations for
temperature and pressure fluctuations. As patients breathe into the spirome-
ter, care must be taken to avoid contamination of the internal part of the
spirometer. Some of these protective measures are disposable mouthpiece
 Respiration Monitors

and disposable patient breathing circuit, bacterial filter, and heated trans-
ducer chamber to prevent condensation.

++e Airflow

Mouthpiece,
filter, and

-
patient circuit

Flow
transducer
-b Amplifier + Signal
processor
Output and
display
-

Figure 25-7. Flow-Sensing Spirometer.

The volume of gas in the lungs is at body temperature and atmospheric

pressure, and is saturated with water vapor (BTPS). In respiratory volume
measurements, in order to compare the volumes to those in blood, correc-
tion must be made to the volume obtained by a spirometer at ambient tem-
perature and atmospheric pressure. The following equation derived from the
gas laws can be used for the correction of the volume to body temperature
of 37OC and a saturated vapor pressure of 47 mmHg.
+
V (BTPS)= Vt X
273 37
[273+t] [Ps-47 1'
PB- P ~ 2 0

where t = temperature of the gas in spirometer in OC,

Vi = volume collected at t OC,
PB = barometric pressure in mmHg, and
PHJO= water vapor pressure (mmHg) of the gas in the spirometer at t OC.
In practice, the temperature and pressure corrections can be obtained from
published tables.

RESPIRATION MONITORS

Clinical bedside monitoring of respiratory function is useful to assess the

need for further respiratory intervention such as the introduction of mechan-
 Biomedical Device Technology: Principles and Design

ical ventilation. It is also a useful tool in evaluating the maturity of the regu-
latory functions of the respiratory system in neonatal development. The
breathing rates as well as the waveform of breathing are the two parameters
to be measured in respiratory monitoring. In addition, the time elapsed of no
breathing, or apnea, is also monitored. Respiration rate for a normal adult
ranges from about 12 to 16 breaths per minute. Breathing rates for neonates
are much higher (about 40 bpm). An apnea alarm is usually set at 20 sec-
onds.
There are a number of methods to obtain the respiratory waveform and
calculate the respiration rate. The common methods are the impedance
method, which measures the electrical impedance across the patient's chest,
and the thermistor method, which detects the airflow in the patient's airway.

Heated Thermistor Method

This method measures the change in temperature of a heated thermistor
placed in the patient's breathing airway. A thermistor is placed in the air path
of the nostril as shown in Figure 25-8 or in a breathing circuit. A current
source produces a constant current to heat the thermistor so that its temper-
ature is above the ambient temperature but less than the body temperature.
When there is no air flowing across the thermistor, the voltage across the
thermistor remains unchanged. During expiration, the warm air from the
lung heats the thermistor. The voltage across the thermistor will then
decrease with its resistance. During inspiration, the colder outside air cools
the thermistor, which causes the voltage across the thermistor to increase.
The variation of voltage across the thermistor will vary with the airflow of
respiration. This voltage is recorded and plotted against time.

Impedance Pneumographic Method

During inspiration, the volume of the thoracic cavity increases and air is
pulled into the lung. The impedance across the chest therefore becomes
higher. During expiration, the chest volume decreases and pushes air out of
the lung. The impedance across the chest therefore becomes lower. In the
impedance pneumographic method, the monitor derives the respiration
waveform and the breathing rate by measuring the change in impedance
between a pair of electrodes applied on the chest of the patient.
To measure the impedance across the chest, a constant current I is
applied across the chest through a pair of electrodes. The voltage Vmeasured
across the electrodes is therefore proportional to the chest impedance Z (V=
I X 2).Figure 25-9 shows the setup to monitor respiration using the imped-
ance pneumographic method.
 Respiration Monitors

Constant
current source
To signal
processor and

Heated I Heated
thermistor thermistor
Figure 25-8. Heated Thermistor Respiration Monitor.

impedance Constant
current source

waveform display
Figure 25-9. Impedance Pneumographic Respiration Monitor.

In order to prevent muscle and nerve stimulation, the injected current

must be small and of high frequency. In general, monitors use frequencies
higher than 25 kHz and current amplitudes below 50 pA.In fact, the output
of the amplifier in Figure 25-9 is an amplitude modulated signal with the car-
rier frequency equal to the frequency of the applied current. The amplitude
variation of the modulated signal is proportional to the impedance change
due to respiration. Figure 25-10 shows the respiration impedance waveform,
the current source waveform, and the waveform of the detected voltage
across the chest.
In most applications, respiration monitors using the impedance methods
employ the same sets of electrodes for ECG monitoring. Note that the
impedance across the electrodes depends on the tissue impedance and the
electrode-skin interface and is on the order of hundreds of ohms or kilo-
ohms. However, the variation of the impedance due to respiration lies in the
 Biomedical Device Technology: Princ$les and Design

- A
3 Chest impedance
-0
(D
a
n,
3
o
n, 2 Voltage across chest electrodes

-
3
u
2
iii

--< 0

b
T ~ m e(sec)

0
t High-frequency constant current

Figure 25-10. Impedance Pneumographic Respiration Monitor Waveforms.

range of 0.1 to 4 ohms. As a result, the signal is very sensitive to electrode

and body movement. In addition, since tissue impedance is frequency-
dependent, any fluctuation in frequency will create a change in the measured
chest impedance. In order to minimize these errors, the applied current must
have constant amplitude and be derived from a very stable frequency source.
Figure 25-11 shows the functional block diagram of a respiration moni-
tor using the impedance method.
The respiration monitor is part of the ECG/respiration monitor using the

To ECG electrodes:
RA, LA, LL, etc

I I I

Constant Output
Differential
current
driver
-+ Lead
selector -+ Buffers
Synchronous
amplifier -b demodulator amplifier

t 4

To waveform
display and
signal processor

Figure 25-11. Block Diagram of Respiration Monitor Using Impedance Method.

 Respiration Monitors

same set of ECG skin electrodes applied to the patient. The lead selector of
the respiration monitor selects a pair of electrodes from the set of ECG elec-
trodes. The high-frequency current (e.g., 50 kHz) flowing through the
patient's chest from one electrode to another creates a voltage of the same
frequency with amplitude equal to the product of the current and impedance
across the chest. This voltage is captured to derive the respiration waveform
and breathing rate. The voltage signal is first buffered so that it will not affect
the ECG part of the monitor. The synchronous demodulator then removes
the high frequency from the measured voltage and recovers the respiration
waveform.
 Chapter 26

MECHANICAL VENTILATORS

OBJECTIVES

Explain the applications of mechanical ventilation.

Differentiate between positive and negative pressure ventilators.
Classify ventilators based on the methods used to terminate inspiration.
Explain the types of breaths delivered by a ventilator, its modes and
submodes, and the ventilation parameters.
List some of the common operator controls, alarm settings, and emer-
gency modes in positive pressure ventilation.
Sketch a block diagram of a positive pressure ventilator and explain the
functions of each block.
Analyze the pneumatic diagram, trace gas flow, and identify basic com-
ponents of a positive pressure ventilator.

CHAPTER CONTENTS

1. Introduction
2. Types of Ventilators
3. Modes of Ventilation
4. Ventilator Parameters and Controls
5. Basic Functional Building Blocks
6. Pneumatic System Diagram
7. Safety Features
 INTRODUCTION

Ventilators provide assistance to patients who cannot breathe on their

own or who require assistance to maintain a sufficient level of ventilation.
Patients may require ventilation support due to illness (e.g., asthma, CNS dis-
order), injuries, congenital defects, postoperative conditions, or the influence
of drugs (e.g., under general anesthesia).
The first mechanical ventilator was developed in 1927 by Philip Drinker
of the United States. It was known as the "iron lung* for treating victims of
poliomyelitis in the early 1950s. The iron lung is basically an airtight metal
chamber enclosing the entire body of the patient except that the head is out-
side the chamber. The chamber is separated from the outside atmosphere by
an air seal around the neck of the patient. To create inspiration, the pressure
inside the metal chamber is reduced to below atmospheric pressure. As the
outside pressure is higher than the chamber pressure, air is drawn into the
patient's lungs through the patient's airway. Expiration is achieved by return-
ing the chamber to atmospheric pressure.
The iron lung is classified as a negative pressure ventilator as the inspi-
ratory phase of the respiratory cycle is created by a negative pressure. Today,
positive pressure ventilators are used to avoid having to enclose the entire
patient's body in a pressure chamber. A positive pressure ventilator uses a
mask, an endotracheal tube, or a tracheostomy tube to connect the machine
to the patient's airway. The lungs are inflated by positive pressure during
inspiration; expiration occurs upon release of the pressure. This chapter
describes the principles of operation, design, and construction of positive
pressure mechanical ventilators.

TYPES OF VENTILATORS

Intermittent positive pressure breathing (IPPB) therapy is a method to

assist spontaneous breathing by applying a positive pressure during inspira-
tion and allowing passive expiration by removal of the inflated pressure. It is
believed by some that IPPB can reduce the work of breathing, promote
bronchopulmonary drainage, and provide more efficient delivery of bron-
chodilator drugs.
Mechanical ventilators overcome respiration deficiencies by assisting
spontaneous breathing or completely taking over the breathing of the
patient. A positive pressure ventilator inflates the lungs by elevating the pres-
sure in the airway to achieve adequate alveolar gas exchange. Expiration is
achieved by opening the airway to the atmosphere so that the gas in the
 Biomedical Device Technology: Princ$les and Design

lungs is passively released to the atmosphere. Portable ventilators are used in

step-down units, extended care facilities, and homes. Home use portable
ventilators are often used 24 hours a day and are set up and operated by the
patient or home caregivers. Portable ventilators should be user-friendly and
are much simpler in operation than critical care ventilators. For some
patients who cannot tolerate the high inspiratory pressure and rapid airflow,
high-frequency ventilators are used. These ventilators cycle at rates much
higher than normal spontaneous respiration but deliver lower inspiratory
volume for each breath.
Mechanical ventilators are classified according to how they terminate the
inspiration phase of the ventilation cycle. Four parameters can be monitored
and used to terminate inspiration: pressure, volume, time, and flow.
1. A pressure cycled ventilator monitors the pressure in the airway and
ends the inspiration when a certain preset pressure is reached.
2. A volume cycled ventilator measures the volume of gas delivered to
the patient and ends the inspiration when a preset volume has been
delivered.
3. A time cycled ventilator tracks the time of inspiration and ends the
inspiration when a preset time is reached.
4. A flow cycled ventilator senses the pressure in the airway and ends
the inspiration when the inspiration flow falls to a preset level.
Figure 26-1 shows typical pressure and flow waveforms of positive pres-
sure mechanical ventilation measured in the airway of a patient. At the start
of the inspiratory phase, the airway/lung pressure increases rapidly until suf-
ficient gas has entered the lungs. The gas flow rate also decreases as the pres-
sure builds up. During the expiratory phase, the airway is opened to the
atmosphere. The alveolar pressure drops and gas is pushed out of the lungs.

Figure 26-1. Ventilation Pressure and Flow Waveforms.

 Mechanical Entilators

MODES OF VENTILATION

When a patient is undergoing positive pressure mechanical ventilation,

there are two types of allowed breath: spontaneous breath and mandatory
breath. In a spontaneous breath, the breathing parameters are determined by
the patient's condition. However, in a mandatory breath, all breathing para-
meters are determined by the machine. A ventilator-initiated mandatory
(VIM) breath is initiated by the ventilator timing circuit, whereas a patient-
initiated mandatory breath (PIM) is initiated by the patient. In a PIM, the
patient attempts to breathe, and the breathing action creates a small negative
pressure in the lungs and airway. Upon detecting this small negative pressure
change, the machine will deliver a positive pressure breath to help the
patient to breathe. A PIM breath is distinguished from a VIM breath by the
slight negative pressure before the onset of the positive inspiratory pressure
(Figures 26-2a and b). The pressure waveform for a spontaneous breath is
characterized by the small negative inspiratory pressure and small positive
expiratory pressure (Figure 26-2c).

Ventilator-initiated mandatory A Patient-initiated mandatory

7
0,
V)
:
V)

:
V)

5
0,

-
73 -u
Y b
0 0
(a) Time (t) (b) Time (t)

Spontaneous

c
0
(c) Time (t)
Figure 26-2. Mandatory and Spontaneous Breaths.

The modes of ventilation specify the characteristics of breaths delivered

by a positive pressure ventilator in response to the patient's breathing
attempts. Some of the common modes of ventilation are briefly explained
below. Their typical pressure waveforms are shown in Figure 26-3.
 Biomedical Device Technology: Princijdes and Design

Controlled mandatory ventilation (CV)-Consists of ventilator-initiat-

ed mandatory (VIM) breaths at prescribed time intervals. The CV
mode is for patients who cannot breathe by themselves.
Intermittent mandatory ventilation (1MV)-Consists of VIM breaths at
prescribed time intervals but allows the patient to breathe sponta-
neously between the controlled breaths.
Continuous mandatory ventilation (CMV)-Consists of a mix of VIM
breaths at prescribed time intervals or mandatory breaths initiated by
the patient (PIM).
Synchronized intermittent mandatory ventilation (S1MV)-produces
VIM breaths at prescribed time intervals if there is no patient breath-
ing initiation. The VIM breaths will be synchronized with the patient's
inspiration effort to generate a PIM breath. The patient is allowed,
within the s-phase window, to breathe spontaneously after a PIM
breath. If the patient initiates a breath after the s-phase, the ventilator
will generate a PIM breath. By decreasing the ventilation rate, carbon
dioxide will build up in the patient's blood, stimulating the respirato-
ry control centers, and trigger the patient to initiate mandatory (PIM)
and spontaneous breaths. This is used to gradually return the work of
breathing to the patient; the process is called "weaning."
Continuous positive-airway pressure (CPAP) ventilation-in the CPAP
mode, all breathing is spontaneous. A continuous positive pressure is
maintained throughout the breathing cycle. It is used to give supple-
mental oxygen and raises baseline pressure to prevent the collapse of
the small airway and alveoli.
In addition to the ventilation modes described above, the following sub-
modes are commonly found in critical care ventilators.
Positive end-expiratory pressure (PEEP)-PEEP is used with manda-
tory breaths; it maintains the lung at a positive pressure at the end of
expiration (Figure 26-3). It is often used to increase the patient's arte-
rial oxygen saturation without increasing the inspired Oz%.
Apnea-Apnea is the cessation of breathing. When the apnea alarm is
activated, the ventilator will measure the duration of no breathing. On
detection of an apnea, the ventilator will automatically go into a pre-
set ventilation pattern stored under the apnea submode.
Pressure support-pressure support creates an elevated baseline pres-
sure during a spontaneous inspiration. By elevating the pressure dur-
ing inspiration, the patient does not have to create the entire pressure
gradient to obtain a meaningful tidal volume. Thus, this mode will
reduce the patient's inspiratory work while still allowing the patient to
control many other breathing parameters.
 Mechanical Kntilators

y I<- Fixed rate --------3

2rn 0
--5
-0
CV

! I b
0
4------ Fixed rate --------3 Time (t)
A
IMV

! u
- I b
A
CMV

Y b

............................,......................... Elevated pressure CPAP

Elevated pressure PEEP

Figure 26-3. Modes of Ventilation.

Sigh-a sigh is a breath delivered by the ventilator that differs in dura-

tion and pressure from a normal breath. The profile of a sigh is often
user-programmable.

VENTILATION PARAMETERS AND CONTROLS

The basic parameters of ventilation are pressure, flow, volume, and time.
These parameters are interrelated during mechanical ventilation.
Pressure-Pressure is the driving force against the resistance of the
patient circuit and airway that causes flow. Unit of measurement is in
 Biomedical Deuice Technology: Princ$les and Design

centimeters of water (cmHzO).

Flow-Flow is the rate of gas at which the tidal volume is delivered. A
higher flow selection will require a higher pressure to create given the
same air resistance in the breathing circuit and patient's airway. A
higher flow will take less time to deliver the required tidal volume.
The unit of flow measurement is liters per minute (LPM).
Volume-Volume measures the quantity of gas. It is usually displayed
as minute volume (total inspired gas volume in 1 minute), which is the
tidal volume times the respiration rate in breaths per minute. The unit
of volume measurement is liters (L) and minute volume is liter-minute.
Time-Time is associated with how frequently a breath is delivered
and also with the time of inspiration and expiration. The unit of mea-
surement for breathing rate is breaths per minute (bpm). A decimal
value is used to indicate the I:E ratib, the ratio of inspiratory to expi-
ratory time in a breathing cycle.
As these parameters are measured outside the patient's body under dif-
ferent conditions than inside the lungs, it is important to correct these para-
meters to body temperature, sea level pressure, and gas saturated with water
vapor (BTPS) before they are used to compare with preset parameters or to
control ventilation.
Operator control parameters (with typical setting ranges) that can modi-
fy the ventilation pattern of a mechanical ventilator include:
Ventilation mode (CV, IMV, SIMV, etc.)
Inspiratory flow waveform (square, descending ramp, etc.)
I:E ratio (0.2-1)
Tidal volume (0.1-2.5 L)
Peak flow (120-180 LPM)
Respiratory rate (0.5-70 bpm)
Sensitivity (0.5-20 cmHzO above PEEP)-for sensing patient-initiated
breathing effort
Inspired air Oz concentration (21-100%)
Manual breath (or sigh) (0.1-2.5 L not exceeding twice tidal volume,
1-3 sighs per hour)
PEEP/CPAP pressure (0-45 cmHzO)
Some common safety alarm features and settings are:
High pressure limit (10-120 crnHzO)
Low inspiratory pressure (3-99 cmHzO)
Low PEEP/CPAP pressure (0-45 cmHzO)
Low exhaled tidal volume (0-2.5 L)
Low exhaled minute volume (0-60 L)
High respiratory rate (0-70 bpm)
 Mechanical Vintilators

Low oxygen/air inlet pressure (35 psig)

Apnea interval (10-60 sec)
I:E ratio (>1)
Some emergency modes of ventilator operation are:
Apnea ventilation-Delivers preset ventilation when apnea is detected.
Backup ventilation-Should the ventilator fail to provide the ventila-
tion to the patient, backup ventilation function will be activated. For
example, activation of backup air compressor to take over failed med-
ical gas supplies.
Safety valve open-If ventilator fails, the patient breathing circuit is
open to the atmosphere to allow spontaneous breathing and manual
bagging.

BASIC FUNCTIONAL BUILDING BLOCKS

Figure 26-4 shows a block diagram of a positive pressure mechanical

ventilator. The following paragraphs describe the functional building blocks.

X A
Patient circult t
Pneumatrc
system
4
0 2
Air

Humidifier}
Monitoring
' 4
nebulizer -

Processor

t
Safety

Figure 26-4. Functional Block Diagram of a Mechanical Ventilator.

Medical aidoxygen supplies-Medical gas from wall outlets provides

the sources of air and oxygen necessary to produce the breathing gas
mixture delivered to the patient. For some ventilators, compressed gas
 Biomedical Device Technology:Principles and Design

cylinders are used as backup gas supplies in case of wall supplies fail-
ure.
Pneumatic system-The pneumatic system regulates the gas pressure,
blends the air and oxygen to desired proportion, and controls the ven-
tilation flow profile according to the control settings.
Patient circuit-It physically connects the pneumatic system to the
patient. It supplies the inspired gas to the patient and removes the
expired gas from the patient. It has one or more check valves to sepa-
rate the inspired and expired gas flow and contains bacteria filters to
prevent contamination.
Processor-According to the user input and the information from the
sensors, the processor produces control signals to the pneumatic cir-
cuit to produce breaths with desired characteristics.
Monitoring-It measures the performance of the pneumatic system
and feeds information back to the processor. Pressure and flow sensors
at different locations of the pneumatic circuit are used to monitor and
control ventilation parameters. Oxygen sensors are used to monitor
the correct air/oxygen mixture being delivered to the patient.
User interface-It allows users to set up ventilation parameters and dis-
plays system and patient information.
Safety/backup-This system protects the patient under ventilation. It
alerts the operator when preset conditions are violated and may initi-
ate backup responses preset by the operator. In case of extreme cir-
cumstances, such as a loss of a gas source, the safety/backup system
may take control of the pneumatic system and override settings previ-
ously selected by the operator.
Humidifier (optional but often required)-Humidifiers are used to
increase the water moisture content in the breathing gas before it is
delivered to the patient. During normal breathing, the inspired gas is
warmed and moisturized as it is passing through the airway. During
mechanical ventilation, prolonged inhalation of dry gas will cause
patient discomfort and may damage the airway tissues. When a
humidifier is used, the inspired gas in the patient circuit is bubbled
through a reservoir of warm water to pick up moisture before entering
the patient's airway. To prevent heat damage to the airway tissues, the
temperature of the inspired gas must be monitored (by a temperature
sensor) to ensure that it is below 42OC.
Nebulizers-Nebulizers are used to deliver medication into the
patient's airway during ventilation. The size of the vapor droplets
determines the site of deposition. Larger droplets deposit in the upper
airway, while tiny droplets (<1 pm) are deposited in the alveoli. Jet
venturi or ultrasound transducers are commonly used to produce tiny
 Mechanical Entilators

droplets of water in the inspired gas. However, a nebulizer has the

potential to deliver too much water and overhydrate the patient.
* Air compressor (optional)-It produces a 50 psig, 21% 0 2 air source. It
is used as a backup to the medical air supply to allow the patient to
breathe in case of medical air supply failure. The compressor will cut
in automatically when it detects a low pressure in the medical air sup-
ply line.

PNEUMATIC SYSTEM DIAGRAM

Figure 26-5 shows a typical pneumatic system diagram of a positive pres-

sure mechanical ventilator together with the patient breathing circuit and the
medical gas supplies.
Medical air from the piped gas wall outlet is connected to the ventilator
via a water trap and coarse filter to remove water condensation and particu-
lates from the hospital's gas supply system. The inlet to the ventilator con-
sists of another filter and a pressure sensor. The pressure sensor will switch
off the air supply line and sound an alarm if the pressure becomes too low
(e.g., <35 psig). A check valve allows the gas to flow in only one direction,
thereby preventing any contamination of the gas supply due to reverse flow.
Medical air, which is usually about 50 psig from the wall outlet, is reduced
to a lower pressure (e.g., 10 psig) by the air regulator before being mixed
with oxygen in the oxygen/air blender. In addition to creating the desired
breathing gas mixture, the flow control within the blender generates the flow
pattern and controls the ventilation rate. A flow sensor (e.g., hot air
anemometer) monitors the volume flow rate of the air supply. (The oxygen
supply line before the blender is identical to that of the air supply line.)
The patient circuit consists of an inspiratory limb and an expiratory limb
connected at a Y-connection. The breathing gas, which contains the desired
proportion of air and oxygen, exits the ventilator via a bacteria filter and
check valve into the inspiratory limb of the breathing circuit. The gas mix-
ture picks up moisture from the humidifier and medication (if needed) from
the nebulizer before entering the patient's lungs. During the inspiration
phase, the exhalation valve is closed to allow the inspired gas to inflate the
lungs. During the expiratory phase, the flow control valve stops the supply
gas flow, the exhalation valve opens to the atmosphere, and the thoracic cav-
ity collapses and forces the gas to exhale from the lungs through the expira-
tory limb of the patient breathing circuit.
As the exhaled gas from the lungs is at body temperature and saturated
with water vapor, water will condense from the gas as its temperature
becomes lowered in the expiratory limb of the breathing circuit. The bacte-
 Biomedical Device Technology: Principles and Design

I Solenoid valve

t------------- Ventitstor -
t Patient Circud --+

Leaend: Q
pressure
& m
bactena
d
check
b
flow
Q = -
pressure breathing m t r o l gas
swtch regulaor filters valve transducer transducer gas llne I~ne

Figure 26-5. Pneumatic System Diagram of Mechanical Ventilator.

ria filter prevents contamination of the ventilator circuit components by the

exhaled gas from the patient. However, a wet filter has a much lower effi-
ciency to remove bacteria from the expired air. To prevent water condensa-
tion at the bacteria filter, a heater is required to warm the expired gas. A col-
lection vial is also connected to the inlet of the expiration compartment of
the ventilator to remove water condensation and sputum from the patient's
exhaled gas before reaching the filter. A check valve is installed in the expi-
ratory limb to prevent reverse gas flow. A flow sensor measures the exhaled
gas flow before it is vented to the atmosphere.
An elevated baseline pressure can be created by imposing a pressure on
the main expiratory gas flow in the expiratory limb of the patient breathing
circuit. Under CPAP or PEEP mode, the CPAP/PEEP controller exerts a
pressure on the valve seat of the exhalation valve. Only exhaled gas with
pressure higher than the CPAP/PEEP pressure can be vented to the atmos-
phere.
 Mechanical Entilators

SAFETY FEATURES

A mechanical ventilator is a critical life-supporting device for a patient

who cannot breathe by himself or herself. Malfunctioning of any part in the
system can threaten the life of the patient. Many safety features are built into
the system. Some of the common safety features are:
Disconnection alarm-A disconnection alarm is an alarm to detect dis-
connection of the breathing circuit from the ventilator or the patient.
Pressure sensors detect a sudden drop in pressure in the patient circuit.
Care must be taken not to lower the pressure limit too much to avoid
false alarm (e.g., due to pressure fluctuation). Too low a setting may
not be sensitive enough to trigger an alarm if the disconnection is at
the distal end of the circuit.
Air leak alarm-Excessive air leak in the system (especially in the
patient breathing circuit) will compromise ventilation. Most ventila-
tors utilize flow sensors to compare the inspired and expired gas flow
volume. For a leak-free system, the volume of inspired gas over a peri-
od of time is equal to that of the expired gas. In the system shown in
Figure 26-5, the inspired gas flow is measured by the flow sensors in
the air and oxygen supply lines. The sum of these two flow sensors
should be the same as the flow measured by the flow sensor in the
expiration circuit. Note that all volumes must be adjusted to BTPS
before making the comparison.
High pressure alarm-A high pressure alarm can alert the user to a
kink or obstruction in the patient breathing circuit. It can also prevent
lung damage from inadvertent high pressure being developed in the
breathing lines.
Loss of power-Although breathing gas is derived from the medical gas
wall outlets, most critical care ventilators employ electronic or micro-
processor circuits for control and alarm. In case of a power loss, the
ventilator should sound an alarm to alert the clinicians. The ventilator
and the patient circuit should be designed such that manual ventilation
can be performed on the patient in a power failure situation. The unit
should have a battery backup memory to store all the machine settings
and data so that it can be powered up immediately without having to
undergo lengthy initialization and programming after power has been
restored.
Loss of gas supplies-Ventilators should be designed to allow automat-
ic switch-over of oxygen to medial air if there is no supply of oxygen
(and vice versa). Similar to a loss of power, the ventilator should allow
manual ventilation when all gases are lost. Some ventilators have a
built-in electrical air compressor to supply compressed air to the gas
 Biomedical Device Technology: Principles and Design

lines in case the hospital gas supply has failed or is not available.
Power-up self-test-Many critical care ventilators have a power-up self-
test to check most operational conditions, inchding electronic diag-
nostic and leakage test of the pneumatic circuit. The compliance of the
system, including the patient breathing circuit, is measured during the
test to ensure accurate calculation of all breathing parameters.
 Chapter 27

ULTRASOUND BLOOD FLOW DETECTORS

Describe the properties of ultrasound.

Define the equations of sound propagation and Doppler effect in ultra-
sound.
State the principles of Doppler and transit time blood flow measure-
ment techniques.
Sketch a block diagram of a Doppler blood flowmeter and describe the
functions of each block.

CHAPTER CONTENTS

1. Introduction
2. Ultrasound Physics
3. Transit Time Flowmeter
4. Doppler Flowmeter
5. Functional Block Diagram of a Doppler Flowmeter

INTRODUCITON

Blood flowmeters and detectors are used to measure and evaluate the
flow of blood in blood vessels. An ultrasound blood flow detector can be
used noninvasively to locate and assess the degree of vascular restriction. For
example, an ultrasound blood flow detector is used to perform postoperative
assessment after vascular surgery or in detecting carotid artery occlusion by
435
 Biomedical Device Technology: Principles and Design

examining the pattern of dominant periorbital collaterals. In addition to

detecting flow, some ultrasound blood flowmeters can noninvasively quanti-
fy the velocity and the volume of blood flow in blood vessels.

ULTRASOUND PHYSICS

Sound is a mechanical longitudinal (or compression) wave in which par-

ticles move back and forth parallel to the direction of wave travel. Ultra-
sound is sound beyond the upper audible frequency limit of human beings
(i.e., of frequency 20 kHz or higher). Low-intensity (e.g., < 0.1 W/cm2)ultra-
sound is absorbed by human tissue without known damage. However, high-
intensity ultrasound (e.g., 500 W/cm2) can cause tissue injury due to heating
effects. In addition, with appropriate frequency and setup, ultrasound can
create shock wave and cavitation in tissues.
The wavelength h of ultrasound is equal to the velocity c of sound in the
medium divided by its frequency or

Listed below are the propagation speeds of ultrasound in different media.

In diagnostic ultrasound, the average velocity of sound in soft tissue is 1,540
m/s or 1.54 mm/ps. This average velocity is used in distance calculation and
in assessing the propagation of sound in body tissue.

Table 27-1.
- -
Propagation
- -
Speed of Ultrasound.
- - - -

-
Medium
- -
Propagation Speed (m/s)-
- - -
Air 300
Water 1,480
Soft tissue 1,440 to 1,640
Fat 1,450
Bone 2,700 to 4,100
~~~~~~

The distance of sound travel is equal to the velocity times the time of travel
or d = vt.
When an ultrasound source and a receiver are moving at velocities of K
and K, respectively, as shown in Figure 27-1, the apparent frequency8 of the
ultrasound signal detected by the receiver is different from the source fre-
quency$ The difference, called the Doppler shift, depends on the source fre-
 - -
Ultrasound Blood Flow Detectors

vs vr
Source Receiver
Figure 27-1. Doppler Effect.

quency as well as the veIocities of the source and the receiver. This is known
as the Doppler effect.
In the case of the arrangement in Figure 27-1, the frequency of the
received ultrasound is:

where V7 = velocity of the receiver away from the source,

K = velocity of the source in the same direction as R,
f= frequency of the source, and
C= speed of sound (in air = 330 m/s).
The Doppler shift is defined as the change in frequency when the source
and receiver are moving relative to each other. The Doppler shift is:

Example
What is the Doppler shift when the source is stationary and the receiver is
moving toward the source at 100 m/s in air? ( C = 330 m/s).

Solution
Using equation (3), substituting C= 330 m/s, R = -100 m/s, and K = 0.0,

TRANSIT TIME FLOWMETER

A transit time flowmeter computes the flow velocity by measuring the

time difference between the sound traveling upstream and downstream of
the flow. Figure 27-2 illustrates the principles of an ultrasound transit time
 Biomedical Device Technology: Principles and Design

flowmeter to measure blood flow in a blood vessel. An ultrasound transmit-

ter and a receiver are positioned at an angle 0 external to the blood vessel.
To start the measurement, the ultrasound transmitter A emits a short pulse.
The time for the sound to reach receiver B is measured. In the next phase,
ultrasound transmitter B emits a sound pulse. The time it takes to reach
receiver A is again recorded. The velocity of blood flow in the vessel
depends on the difference between the two recorded times. Below is the
derivation:

UIS transmitter and receiver

\
\

D
\

\ ,
//
' 7-
B UIS transmitter
and receiver

,
"/

,,
Figure 27-2. Transit Time Flowmeter.

In the downstream transmission, the time for the sound to travel from
point A to B is:

TAB
= (downstream),
C + vcose
where D = the distance between the ultrasound transducers,
C= the velocity of sound in the medium,
13 = the angle between the direction of sound travel with the direction of
blood flow, and
v = the velocity of blood.
In the upstream transmission, the time for the sound to transmit from
point B to A

TBA=
c - V C O S ~(upstream),
 Ultrasound Blood Flow Detectors

The time difference AT between the upstream and downstream trans-

mission is:

AT = TBA- TAB
= D - D - 2DvGsO
C - vcose C + vcose - C' - v 2 C 0 ~ 0

DOPPLER FLOWMETER

Doppler flowmeters make use of the Doppler effect to determine the

velocity of flow. For the setup shown in Figure 27-3, instead of a moving
source and a moving receiver as shown in Figure 27-1, both transmitter and
receiver are stationary. The sound wave is reflected from a moving reflector
traveling at the same speed as the fluid flow. In blood flow measurement, the

U/S transmitter

*4?-

Q
& UIS receiver
B
Figure 27-3. Doppler Flowmeter.

fluid is blood and the reflector is a red blood cell.

From the doppler shift equation (equation 3)
 Biomedical Device Technology: Principles and Design

C VCos4 +
Note that the negative sign indicates a decrease in frequency.

If 0 and 4 are both zero, f~ = -2 v fs.

-
C

Example
For the ultrasound Doppler blood flowmeter as shown in Figure 27-3, if 0 =
= 60°, V= 100 cm/s, $= 5 MHz, and C = 1.5 x 1 0 cm/s, what is the mag-
nitude of the Doppler shift?

Solution

f~ = 5 X 10" X loo
1.5 X 1 0
X (Cos 60' + Cos 60') Hz = 3.3 kHz.

Note: The above results showed a single frequency shift. In a real situation,
as blood cells travel at different velocities, the backscattered ultra-
sound received will be of a broad frequency range.

In practice, an ultrasound Doppler blood flowmeter has the transmitter

and receiver together so that the probe (containing both the transmitter and
receiver) can be placed on the surface of the skin or on top of a blood vessel
during blood flow velocity measurements. The Doppler shift in this case
becomes:

Ultrasound probe UIS transmitter and receiver

Red blood cell (reflector)

Figure 27-4. Doppler Blood Flowmeter.

 Ultrasound Blood Flow Detectors

fD =- 2 vfs.
C
FUNCTIONAL BLOCK DIAGRAM OF
A DOPPLER BLOOD FLOWMETER

,
Radio Audio
frequency + Detector + High pass
filter + frequency audio
speaker
amplifier amplifier
I

Ultrasound
p f l transmine[
I
I
I
1 Zero
crossing
detector 1
Radio
frequency To output
oscillator integrator recorder

Figure 27-5. Doppler Blood Flowmeter Block Diagram.

Figure 27-5 shows a functional block diagram of an ultrasound Doppler

blood flowmeter.
The radio frequency oscillator generates the RF (e.g., 5 MHz) excitation
signal to the ultrasound transmitter. The receiver detects the ultrasound
reflected from the moving red blood cells in the blood vessel. The RF fre-
quency and the Doppler angle can be chosen such that the Doppler shifts
due to the traveling blood cells are in the audio frequency range (see pre-
ceding example). If all the blood cells are moving at one constant velocity,
the frequency of the received signal will have only one frequency, which is
equal to the transmitter frequency plus the Doppler shift @ + Ji). However,
as blood flow is pulsatile and blood flow velocity is not the same across the
blood vessel, J% is not a single value and will occupy a range of frequency.
The signal received is a frequency-modulated signal with the Doppler shift
proportional to the blood flow velocity. The detector is an FM demodulator
that removes the transmitter frequency ~5from the signal. In most cases, the
output also contains a large amplitude low-frequency wave caused by the
blood vessel-wall-motion. This vessel-wall-motion artifact can easily be
removed by a high pass filter. An audio frequency amplifier intensifies this
signal and sends it to an audio speaker. Figure 27-6a shows the output from
 Biomedical Device Technology: Princ$les and Design

the detector and filter. As the Doppler shift is in the audio frequency range,
the clinician can hear the flow pattern of the blood in the blood vessel. A
high pitch (large Doppler shift) corresponds to fast-moving blood and a low
pitch corresponds to low blood flow. The Doppler shift (which is propor-
tional to the blood flow velocity) can be converted to an analog flow veloci-

(c) Time ( s )
Figure 27-6. Doppler Flow Detector Waveforms.
 Chapter 28

FETAL MONITORS

OBJECTIVES

Describe the clinical significance of fetal heart rate and maternal uter-
ine activities during labor.
Describe and contrast different methods of monitoring fetal heart rates,
including direct, ultrasonic, maternal abdominal, and phono methods.
Describe and compare external and intrauterine methods of monitoring
uterine activities.
Explain the construction and principle of transducers and sensors used
in fetal monitoring.
Sketch a simple block diagram of a fetal monitor.

CHAPTER CONTENTS

1. Introduction
2. Monitoring Parameters
3. Methods of Monitoring Fetal Heart Rate
4. Methods of Monitoring Uterine Activities

Electronic fetal monitoring provides graphical and numerical informa-

tion to assist the clinician to assess the well-being of the fetus. During labor,
the fetal heart rate often accelerates and decelerates in response to the uter-
ine contractions and fetal movements. Characteristics of these patterns may
 Biomedical Device Technology: Principles and Design

reveal labor problems such as fetal hypoxia or decreased placental blood

flow. Examining these patterns may indicate alternative courses of labor (e.g,
cesarean section or forceps delivery) or drug therapy (e.g., administering
labor-inducing or labor-prohibiting drugs).
Antepartum (before birth) monitoring is used to monitor the develop-
ment of the fetus in the uterus. Intraparturn monitoring includes monitoring
the status of the mother and fetus as well as the progress of labor. Maternal
monitoring includes measurements of the mother's vital signs such as heart
rate, respiratory rate, blood pressure, temperature, oxygen saturation level,
and uterine activity. Fetal monitoring refers to the monitoring of the fetal
heart rate (FHR) and the maternal uterine activity (UA) during labor and
delivery. Fetal monitors were first available in the late 1960s. Today, fetal
monitoring is used in more than 60% of deliveries in North America.

MONITORING PARAMETERS

The two primary parameters in fetal monitoring are fetal heart rate
(FHR) and uterine activities (UA). Other parameters that may be monitored
are the maternal ECG and 010Sa0~Normal fetal heart rates fall within the
range of 120 to 160 beats per minute (bpm) during the third trimester of
pregnancy and fluctuate from the baseline rate during contractions. Fetal
heart rate may reveal the conditions of the fetus during labor and delivery.
Figure 28-1 shows a typical recording of FHR. Some abnormal FHR con-
ditions and their indications are:
Tachycardia (high heart rates)-may be caused by maternal fever, fetal
hypoxia, immaturity of fetus, anemia, or hypotension
Bradycardia (low heart rates)-may be caused by congenital heart
lesions or hypoxia
Variation-too much fluctuation indicates stress or hypoxia

Figure 28-1. Fetal Heart Rate.

 Fetal Monitors

Uterine activity refers to the frequency and intensity of the contractions

of the uterus. During labor, the smooth muscles of the uterus contract rhyth-
mically, thereby increasing the pressure of the amniotic fluid and forcing the
fetus against the cervix. UA indicates the progress of labor. Figure 28-2
shows a typical recording of UA.

"

F i p e 28-2. Uterine Activities.

The usual characteristics of uterine activities are:

Frequency (F)-less than once in 3 minutes is slow progress of labor
Duration (T)-under 45 seconds of contraction is slow progress of
labor
Amplitude (A)-above 75 mmHg usually indicates active labor
Shape-the shape of the contraction pressure is normally bell-shaped.
An irregular shape may indicate labor pushing, fetal movement,
maternal respiration, or blocked catheter
Rhythm-couplets and triplets indicate abnormal activities
Resting tone (pressure between contraction)-about 5 mmHg for non-
labor and rise to 20 mmHg for induced labor

METHODS OF MONITORING FETAL HEART RATE (FHR)

Fetal heart rate may be obtained by listening to the heart sound of the
fetus, directly connecting electrodes to the fetus, applying electrodes on the
abdomen of the mother, or using Doppler ultrasound. The three methods are
described in the following sections.
 Biomedical Device Technology: Princ$les and Design

Direct ECG (DECG)

Direct ECG is an invasive method that connects a spiral electrode to the
scalp of the fetus. During application, the electrode is inserted through the
vulva. While pushing against the scalp of the fetus, the clinician applies a 360
degree turn to the spiral electrode so that the electrode is screwed and
secured into the skin of the scalp (Figure 28-3a). The electrode can be
applied only when the head of the fetus is accessible; that is, only after the
amniotic sac has ruptured. The other electrode is usually a skin electrode
applied to the thigh of the mother. As the procedure is invasive, it may cause
complications (e.g., infection) to the fetus.

Phono Method
The F H R may be derived by listening to the fetal heart sound. Although
a microphone can be used, this is usually done manually by the obstetric
nurse or physician using a stethoscope placed on the abdomen of the moth-
er. The weak fetal heart sound is usually buried among the louder maternal
heart sound and other sounds (such as sound from bowel movement) within
the mother's body. The advantage of this method is that it is noninvasive and
does not require expensive equipment.

Abdominal ECG
Abdominal ECG is obtained by applying skin electrodes on the
abdomen of the mother (on fundus, pubic syrnphysis, and maternal thigh).
The electrodes ar e attached to a normal ECG machine so that the waveform
and heart rate are displayed. As the electrodes will inevitably pick up the
maternal ECG, careful electrode positioning to capture the fetal ECG and
differentiate them from the maternal signal is required.

Ultrasound Method
Another noninvasive method to monitor FHR employs a Doppler ultra-
sound detector. A beam of continuous wave ultrasound (e.g., 2 MHz) from
an ultrasound transmittedreceiver pair is applied to the abdomen of the
mother (Figure 28-3b). If the ultrasound beam crosses the fetal heart, the
Doppler shift detected from the reflected sound will record the motion of the
fetal heart wall and thus can be processed to obtain the FHR (see principle
of Doppler blood flow detector in Chapter 27). This method provides an
accurate beat-to-beat measurement of the heart rate provided that the ultra-
 Fetal Monitors

sound beam covers the fetal heart. To avoid picking up movement artifacts
from other organs, a narrow sound beam is preferred. However, with a nar-
rower sound beam, the transducer position must be checked from time to
time to ensure that the sound beam is focused on the fetal heart. In addition,
it requires good skin-transducer contact (achieved by application of ultra-
sound gel) to obtain good signal. Although more complicated and expensive,
a pulsed Doppler with time gating can provide better quality signal than a
continuous wave Doppler unit.

ECG monitor Doppler

(a) Direct ECG (b) Ultrasound Doppler Heart Rate Detector

Figure 28-3. Fetal Heart Rate Monitors.

METHODS OF MONITORING UTERINE ACTIVITIES (UA)

Intrapartum uterine activities may be obtained by using an external pres-

sure transducer applied on the abdomen of the mother, or by inserting a
fluid-filled catheter into the uterus. The former is an indirect and noninva-
sive method, while the latter is direct and invasive.

External Pressure Transducer Method

By placing a pressure transducer on the abdomen close to the fundus and
secured by a belt wrapped around the abdomen (Figure 28-4a), the pressure
in the uterus during contractions can be monitored. The pressure exerted on
the transducer varies roughly in proportion to the strength of the contraction.
The transducer is often referred to as the TOCO transducer. The advantage
of this method is its noninvasiveness. However, it has low accuracy (about
20% error) and it requires frequent repositioning and retightening of the belt.
 Biomedical Device Technology: Principles and Desip

Intrauterine Pressure (IUP) Method

The pressure obtained in this method is more accurate than using the
TOCO transducer. It is a direct pressure measurement method using a setup
similar to direct blood pressure monitoring. A fluid-filled catheter is inserted
into the uterus after the amniotic sac is ruptured. The catheter is connected
to a disposable (or reusable) pressure transducer (Figure 28-4b). The pres-
sure inside the uterus is displayed on a blood pressure monitor. Although this
method is more accurate, it is invasive. Care should be taken to ensure that
there is no obstruction or occlusion of the catheter during labor and that the
transducer and setup are properly zeroed before use (see Chapter 18 on
blood pressure monitors for reasons and methods of pressure transducer
zeroing).

c Catheter
(a) External Pressure Transducer (b) lntrauter~nePressure Catheter

Figure 28-4. Uterine Activity Monitoring.

 Chapter 29

INFANT INCUBATORS, WARMERS, AND

PHOTOTHERAPY LIGHTS

OBJECTIVES

Describe the clinical functions of infant incubators and warmers.

List typical features of an infant incubator.
Sketch a functional block diagram of a typical infant incubator.
Explain the construction and major components of an infant incubator.
Identify potential hazards associated with infant incubators.
Describe the mechanism of phototherapy and its clinical functions.
Identify the light spectrum and intensity for effective phototherapy.
State functional features and parameters of phototherapy light source.
Analyze factors affecting effective output of phototherapy light.

CHAPTER CONTENTS

1. Introduction
2. Purpose
3. Principles of Operation
4. Potential Safety Hazards
5. Functional Components and Common Features
6. Phototherapy Lights
 Biomedical Device Technology: Principles and Design

An infant incubator provides a controlled environment to warm the

infant by regulating the air temperature within the incubator chamber. In
additional to air temperature, the humidity and oxygen content within the
chamber can be regulated. A phototherapy light is used to break down
excessive concentration of bilirubin in the newborn. These devices are found
in neonatal care units to treat preterm or sick infants.

PURPOSE

At birth, an infant's body temperature tends to drop significantly due to

heat loss from the body. Heat loss can be through conduction (contact with
other objects), convection (heat carry away by air circulation), radiation (heat
loss to cooler environment due to infrared radiation from the warm body),
and evaporation (latent heat loss from the lungs and skin surface). Most term
neonates regulate their body temperature naturally to some extent.
However, preterm neonates, with thinner skin and higher surface to volume
ratio, tend to lose more heat and can easily become hypothermic. Infant
incubators and radiant warmers are used to reduce heat loss. An infant
warmer radiates heat to the infant by using an external heat lamp (e.g., a
quartz bulb) directed to the infant. Incubators usually have better tempera-
ture regulation than infant warmers and provide a controlled environment
where infants can receive the therapies they need. However, the enclosed
chamber of an incubator is less convenient for accessing the infant than the
open design of an infant warmer.

PRINCIPLES OF OPERATION

An infant incubator consists of an infant compartment enclosed by a

clear plastic hood. The infant lies on the mattress inside the infant compart-
ment. Access doors and ports through the hood allow relatively easy access
to the infant for feeding, examination, and treatment. A blower and heater
underneath the mattress provide forced circulation of warm air inside the
compartment to warm the infant. An infant incubator usually has two modes
of temperature control: skin temperature and air temperature. In the skin
temperature mode, a temperature senor is attached to the skin of the infant.
The skin temperature signal is compared to the preset value to turn the
 Infant Incubators, Warmers, and Phototherapy Lights

heater on and off. In the air temperature mode, a temperature sensor is locat-
ed inside the hood of the incubator to measure the air temperature. This
measured value is compared to the preset value to turn the heater on or off.
To provide better temperature regulation, proportional heating control
instead of a simple on-off control is used.
In a proportional heating control circuit, the heater can be turned on at
lower than the maximum power rating. During initial startup, when there is
a large difference (AT) between the measured and set temperatures, the
heater is switched on at full power. When the difference becomes smaller, the
heater will run at a lower power setting. This control approach reduces the
fluctuations of temperature within the incubator compartment. Figure 29-1
shows an example of the power and temperature relationships of a four-level
proportional heater controller of an infant incubator. When the temperature
difference AT is larger than 6OC, the heater is running at 10O0Iopower, as the
air inside the incubator becomes warmer, the power of the heater is reduced.
When the temperature inside the incubator is less than the preset tempera-
ture by less then 2OC, the heater is running at only 25%.

0 2 4 6 8
Figure 29-1. Proportional Heater Control Characteristics.

Proportional heating control can be implemented by using several banks

of heaters (e.g., using four 250 W heaters instead of one 1,000 W heater) or,
if a single heater is used, by varying the duty cycle of the heater supply
power. The latter can be designed to produce a continuous variation of
heater power according to the measured temperature difference.
Most incubators allow users to vary the relative humidity inside the units.
Humidity can be controlled by adjusting the amount of airflow through a
reservoir of water underneath the mattress or by using an external humidifi-
 Biomedical Device Technology: Princ@les and Design

er. Most incubators have an oxygen inlet to create an elevated oxygen level
within the incubator. Some have a built-in oxygen controller to maintain an
elevated level of oxygen inside the chamber.

POTENTIAL SAFETY HAZARDS

In most cases, infants inside incubators are premature with poor regula-
tion. Deaths and injuries to infants have been linked to temperature regula-
tion failures causing incubator overheating and infant hyperthermia. A
detached skin sensor under the skin temperature control mode detects air
instead of skin temperature, which will lead to overheating. Periodic check-
ing of heat sensor mode, temperature setting, and sensor condition (proper
attachment of skin sensor) is highly recommended.
Poor oxygen control may cause hyperoxia or hypoxia. Excessive high
oxygen concentration can lead to retrolental fibroplasia (formation of fibrous
tissue behind the lens) in premature infants.
As the infant must stay inside the incubator around-the-clock, it is impor-
tant to reduce the noise level inside the hood to protect the hearing of the
infant. The main source of continuous noise is the motor and the blower.
Most manufacturers are able to reduce the noise level to below 50 dB.
However, using a nebulizer, opening and closing of access doors, et cetera
can produce a temporary noise level as high as 100 dB.
The warm and moist environment inside the incubator and the water
reservoir for humidification can become an incubator for bacteria. Care must
be taken to clean and disinfect incubators after every use. Complete disin-
fection or sterilization should be done periodically.

FUNCI'IONAL COMPONENTS AND COMMON FEATURES

Figure 29-2 shows the functional component diagram of an infant incu-

bator. Common features of infant incubators are:
Easy access-front, side, and rear access with cuffed ports to minimize
temperature fluctuation
Adjustable mattress elevator and tilt mechanism
Skin or air temperature sensors for temperature control
Proportional heating control to maintain temperature and minimize
temperature fluctuation
Oxygen sensor, regulator, and supply manifold to control oxygen
level
 Infant Incubators, Warmers, and Phototherapy Lights

n 0 2 sensor Skin temperature sensor Air temoerature sensor

level
0 2
control control control

with filter

Figure 29-2. Functional Component Diagram of an Infant Incubator.

Humidity sensor and water reservoir to maintain relative humidity

inside incubator
Low airflow across infant to reduce heat loss and dehydration
Low internal audible noise to minimize noise and prevent hearing
damage to infant
Alarms-including temperature, overheat, oxygen level, humidity,
heater, and loss of airflow
Independent maximum air temperature (41°C) sensor and alarm to
prevent overtemperature
Display-numerical display including temperature, oxygen level,
humidity level, and heater power
Data trending and alarm log
Air and oxygen inlet filters
Easy to clean and disinfect

PHOTOTHERAPY LIGHTS

A phototherapy light is used to break down bilirubin in the newborn.

Jaundice occurs when the liver of the infant has not reached full detoxifica-
tion capability, especially in premature infants. During the first week of life,
infants have poor liver function to remove bilirubin. A bilirubin level of 1 to
 Biomedical Device Technology: Principles and Design

5 mg per 100 ml of blood within the first 3 days of birth is considered nor-
mal. This level should decrease as the liver begins to mature. Visible light
spectrum of wavelength from 400 to 500 nm (blue) has been shown to be
effective in transforming bilirubin into a water-soluble substance that can
then be removed by the gallbladder and kidneys. A spectral irradiance of 4
pW/cmVnm at the skin surface is considered to be the minimum level to pro-
duce effective phototherapy.
A phototherapy light can be placed directly over an infant in a bassinet
or placed over the hood of an incubator. Blue light sources are used to
increase the efficacy of phototherapy. However, blue light can mask the skin
tone and is hard on the eyes of the caregivers. As a compromise, some man-
ufacturers may use a combination of white and blue light sources and have
built-in features to switch off the blue lights during observation.
Ultraviolet radiation (<400 nm) emitted from most blue light sources is
harmful to the infant. Infants receiving phototherapy are required to wear
eye protectors to prevent damage to their retinas. A PlexiglasTM (used as an
enclosure) placed between the light source and the infant can cut out most of
the wavelength below 380 nm. Far-infrared radiation (heat) can create hyper-
thermia and cause dehydration to the infant. As a precaution, monitoring or
periodically checking the skin temperature of an infant receiving photother-
apy treatment is recommended.
Some common features of a phototherapy light are:
A blue light source (e.g., special blue fluorescent tubes, tungsten halo-
gen, etc.) with a high-intensity blue spectrum (e.g., 400-500 nm) or a
combination of blue and white light sources is used.
The output of phototherapy lights in the range of wavelength from 400
to 500 nm measured at skin level should be greater then 4
pW/cmVnm. Most devices on the market have output much greater
than this minimum level (e.g., 15 pW/cm2).
Filters (PlexiglasTM)to remove ultraviolet (280-400 nm) radiation to
avoid damaging the infant's eyes and skin.
Equipped with white light for observation (with blue light switched off
during observation)
Observation timer to automatically switch back to phototherapy after
observation
Light bulb operation timer to signal light bulb replacement
Light source housing on height adjustable stand
Many manufacturers are using a number (e.g., eight) of 50 cm (20 inch-
es) fluorescent tubes in a metal housing for phototherapy units. The lower
surface of the housing is a piece of PlexiglasTMfor mechanical barrier as well
as serving as an UV filter. A typical unit is shown in Figure 29-3. The light
source housing is either mounted on a height adjustable stand or placed
 Infant Incubators, Warmers, and Phototherapy Lights

Plexiglas filter
Fluorescent tubes
\
Metal housing

Figure 29-3. Gross-SectionatView of a Phototherapy Light.

directly on top of the hood of an infant incubator.

Experiments have shown that the spectral output of a fluorescent tube
changes with the tube surface temperature as well as with time. An enclosed
tube compartment as shown in Figure 29-3 can reach a temperature of 70°C
after 2 to 3 hours of use. Figure 29-4 shows the typical characteristics of a
phototherapy light output with respect to temperature inside the light source
compartment. It shows that the light output will reach a peak value shortly
after it is turned on and will drop to about 70010 of its peak value after a few
hours when it has risen to a steady temperature.
In addition, from manufacturers' specifications, the output of a fluores-
cent tube has a limited life span ranging from a few hundred hours to about

I b
I I
20 40 60 Temp ( O C )
Figure 29-4. Phototherapy Light Output vs. Temperature Characteristics.
 Biomedical Device Technology: Princ$les and Design

2,000 hours. The tube output decreases as it is being used (e.g., 10% drop
after 300 hours). Blue fluorescent tubes generally have a shorter life span
than ordinary white tubes. In order to ensure sufficient light output for pho-
totherapy, some sites are measuring the output using a special light meter
and replacing them when they are below a certain limit. Instead of perform-
ing periodic output measurements, some users implement a fmed schedule
(e.g., monitoring the hours of operation) to replace these light sources.
 Chapter 30

BODY TEMPERATURE MONITORS

Differentiate between core and peripheral temperature and list the sites
for body temperature measurement.
Differentiate between continuous and intermittent temperature moni-
toring.
Describe the principles of operation of a typical bedside continuous
temperature monitor.
Analyze the transducer circuit diagram and the functional block dia-
gram of a typical continuous temperature monitor.
Describe the principles of operation of I R thermometry.
Define the terms emissivity and jeld of view, and explain their signifi-
cance in IR thermometry.
Analyze the functional building blocks of a typical tympanic (ear) ther-
mometer.
State the sources of error in body temperature measurement using tym-
panic thermometers.

CHAPTER CONTENTS

1. Introduction
2. Sites of Body Temperature Measurement
3. Continuous Temperature Monitors
4. Block Diagram of a Continuous Temperature Monitor
5. Infrared Thermometers
6. Theory of Infrared Thermometry
 Biomedical Device Technology: Princ$les and Design

7. Tympanic (Ear) Thermometers

8. Block Diagram of a Tympanic Thermometer
9. Potential Errors in Tympanic Thermometers

INTRODUCTION

The body temperature of a healthy person is regulated within a narrow

range despite variation in environmental conditions and physical activity.
Illness is often associated with disturbance of body temperature regulation
leading to abnormal elevation of body temperature, or fever. Fever is such a
sensitive and reliable indicator of the presence of disease that thermometry
is probably the most common clinical procedure in use. Body temperature is
also measured during many clinical procedures such as surgery, postanes-
thesia recovery, treatment of hypedhypothermic conditions, et cetera.
A temperature monitor allows measurement and display of a patient's
body temperature, sounds an alarm if it is above or below some preset lim-
its, and tracks its variation over a period of time. Most monitors accept dif-
ferent sensors or probes to measure temperature at different body sites (e.g.,
oral, esophageal, rectal, etc.).
Temperature measurement can be continuous or intermittent. A contin-
uous temperature monitor uses a sensor to acquire the temperature at the
measurement site continuously. An example is skin temperature measure-
ment of an infant in an infant incubator. A temperature sensor is placed on
the infant's skin surface to continuously measure and display the infant's
temperature and, using the measurand to control the heater inside the incu-
bator, to achieve temperature regulation. An oral mercury-in-glass ther-
mometer, which takes about 1 or 2 minutes to obtain a temperature reading
between measurements, is an example of an intermittent temperature mea-
surement device. This chapter describes two temperature measurement
devices-a continuous temperature monitor that uses a contact temperature
sensor; and a tympanic thermometer, which is a noncontact intermittent
temperature monitor that senses the heat being radiated from the patient.

SITES OF BODY TEMPERATURE MEASUREMENT

Heat transfer within the body depends on conduction (heat transfer

between adjacent organs) and convection (heat transfer through movement
of body fluid). One can simply visualize the body as a central core at uniform
temperature surrounded by an insulating shell. In body temperature mea-
 Body Temperature Monitors

surement, it is often desirable to measure the core temperature as it reflects

the true temperature of the internal parts of the body (skull, thorax, and
abdomen). Peripheral or shell (skin and subcutaneous fat) temperature is
often affected by the external environment.
It is not practical to define an exact upper level of normal body temper-
ature because there are variations among normal persons as well as consid-
erable fluctuations in a given individual. However, an oral temperature
above 37.2OC in a person at rest is a reasonable indication of disease. Table
30-1 lists the range of temperature (in degree Celsius) measured from differ-
ent sites on normal individuals of different age groups.
Body sites such as the tympanic membrane, rectum, esophagus, naso-
pharynx, bladder, and pulmonary artery are used for core temperature mea-
surement. In a normothermic patient, these sites yield very similar tempera-
ture values. However, under hypothermic conditions, the rectal site can be
cooler than the others by 1 to 2OC.

Table 30-1.
Normal Temperature Comparisons.
Core Oral Rectal
Adult 36.5-37.6 36.0-37.2 36.3-37.6
Age 7-14 years 36.8-37.3 36.4-36.9 36.4-36.9
Age 3-6 years 37.3-37.6 36.9-37.2 36.9-37.2

CONTINUOUS TEMPERATURE MONITORS

Continuous temperature monitors are used in situations in which a

patient's body temperature is required to be measured continuously.
Examples of such situations are patients under general anesthesia, or receiv-
ing hypothermic or hyperthermic treatment. Many temperature transducers
can be used in continuous temperature monitoring. However, the most com-
mon type of transducer for clinical application is a thermistor. The YSI series
probes (patented by Yellow Spring Instruments) offer true interchangeability
between probes and monitors from different manufacturers without the need
for recalibration or adjustment. The temperature transducer element in a
YSI 400 probe is a precision thermistor manufactured to achieve the resis-
tance-temperature characteristics according to the YSI 400 specifications.
The characteristics are shown in Table 30-2 with less than 0.1010 resistance
tolerance between 0 and 80°C.
 Biomedical Device Technology: Principles and Design

Table 30-2.
-
YSI 400 Resistance/Temperature
- - - - - -
Characteristics.
- - - - .

Res. R Emp OC Res. 0 Temp C

' Res. R Te@ OC Res. R
- - - - -- - - -- - -

75.79 k 24 2,354 40 1,200 80 283.1

54.66 k 25 2,253 41 1,153 85 241.3
39.86 k 26 2,156 42 1,108 90 206.5
29.38 k 27 2,065 43 1,065 95 177.5
21.87 k 28 1,977 44 1,024 100 153.2
16.43 k 29 1,894 45 984.2 105 132.7
12.46 k 30 1,815 46 946.6 110 115.4
9,534 31 1,740 47 910.6 115 100.6
7,355 32 1,668 48 876.2 120 88.1
5,720 33 1,599 49 843.2 125 77.4
4,483 34 1,534 50 811.7 130 68.2
3,539 35 1,471 55 672.9 135 60.2
2,814 36 1,412 60 560.7 140 53.4
2,690 37 1,355 65 469.4 145 47.4
2,572 38 1,301 70 394.9 150 42.3
2,460 39 h249 75 333.5
- -- - - -. - -

The normal output characteristics of a YSI 400 temperature probe is

illustrated in Figure 30-1. It has a negative temperature coefficient and a
highly nonlinear resistance-temperature relationship.
Often, it is desirable to have linear output characteristics. A YSI 700
series temperature probe is manufactured by combining two thermistor ele-
ments (Figure 30-2a) of different temperature characteristics to form a tem-
perature transducer that can be wired to produce linear output characteris-

Temperature (OC)
Figure 30-1. YSI 400 Series Temperature Probe Output Characteristics.
 Body Temperature Monitors

tics. The construction of a YSI 700 series thermistor and its application cir-
cuit is shown in Figure 30-2b. Figure 30-3 shows the equivalent resistance-
temperature relationship of the circuit.

Figure 30-3. Linearized Characteristic of YSI 400 Series Thermistor Circuit.

Many continuous temperature monitors on the market can accept YSI

400 or YSI 700 temperature probes. As there are two lead wires for a YSI
400 series transducer, l/4 inch mono phono jacks are often used as connec-
tors with YSI 400 probes. Whereas 1/4 inch stereo phono jacks (3 wires) are
used with YSI 700 probes.
A YSI 400 temperature probe can be connected to one arm of a
Wheatstone bridge as shown in Figure 30-4. The output voltage Vo will vary
according to the change in resistance RT of the probe. As the bridge output
 Biomedical Device Technology: A.inc$les and Design

has a nonlinear relationship with the resistance RT, the resistance values of
R a and Rb can be chosen such that the output voltage/temperature follows a
piecewise-linear relationship.

Figure 30-4 L. Wheatstone Bridge Temperature Monitor.

In a digital thermometer, the resistance-temperature relationship is

stored in digital memory in a lookup table. Once the resistance of the probe
is measured, the corresponding temperature is determined from the lookup
table. No linearization circuit is required.

BLOCK DIAGRAM OF A CONTINUOUS

TEMPERATURE MONITOR

Figure 30-5 shows a simple functional block diagram of a continuous

temperature monitor. It consists of an excitation circuit to provide either a
voltage or a current source to convert the change of resistance (due to change
in temperature) to a voltage output. This output voltage is amplified, filtered,
and sent to the processor and display modules. The purpose of signal isola-
tion is for electrical safety.

INFRARED THERMOMETERS

Mercury-in-glass and electronic thermometers measure temperature by

placing the probes under the tongue, in the rectum, or under the armpit.
Temperature obtained from these sites may not reflect the true core temper-
 Body Temperature Monitors

Temperature
transducer

Signal Amplifier To processor

Excitation
isolation and filter and display

Figure 30-5. Frontend Block Diagram of Temperature Monitor.

ature as they are subject to thermal artifacts. Infrared (IR) ear thermometers,
also known as ear thermometers or tympanic thermometers, allow users to
measure temperature by inserting a probe into the patient's ear canal. These
thermometers are quick and noninvasive. They can measure temperature
without touching the mucous membrane and can be used on both conscious
and unconscious patients. I R thermometers provide a convenient and fast
alternative to other intermittent temperature measurement methods.

THEORY OF INFRARED THERMOMETRY

I R thermometry or pyrometry has long been used as a noncontact

method to measure temperature in the industry. For example, in foundries,
noncontact I R thermometers are used to measure the temperature of molten
metals from a distance.
IR thermometry relies on the principle that radiation is emitted by all
objects having a temperature greater than absolute zero (0 Kelvin) and that
the emission increases as the object becomes hotter. The temperature of the
object can be determined from the emission spectrum of the radiation in the
infrared region. Devices based on IR thermometry are based on Planck's law
and Wien's displacement law.
Planck's law states that if the intensity of the radiated energy is plotted as
a function of wavelength (Figure 30-6), the area under the curve represents
the total energy radiated at the associated temperature.
Wien's displacement law states that the wavelength Amax corresponding
to the maximum energy intensity in the radiated energy spectrum is given by
the equation:

Am, =
2.89 X lo3
T Pm?

where Tis the object temperature in Kelvin.

 Biomedical Device Technology: Princ$les and Design

For example, if the object is at temperature 37OC,

T = 37 + 273 = 310 K + hmax = 9.32 pm.

I00 Wavelength (pm)

Figure 30-6. Radiated Energy Spectrum of an Object.

Another parameter important to I R thermometry is emissivity.

Emissivity (E) is defined as the ratio of the energy radiated by an object at a
given temperature to the energy emitted by a perfect radiator, or a black-
body at the same temperature. Therefore, the emissivity of a blackbody is
1.0, whereas the emissivities for all "nonblack" objects lie between 0.0 and
1.0. The emissivity of body tissue is about 0.95. For most substances, the
emissivity is wavelength-dependent except for a graybody. A graybody is
defined as an object whose emissivity is the same for all wavelengths. As
most objects are not close to being a blackbody, variation of emissivity with-
out proper compensation will lead to error in I R measurement. Most I R
thermometers restrict the sensing wavelength to a chosen narrow band so
that near-graybody characteristics can be obtained.
In I R thermometry, a sensor is used to collect the radiant energy coming
from the object. Therefore, it is important that the sensor's field of view
(FOV) be aligned properly with the object to be measured. The FOV is the
angle of vision at which the instrument operates. It is determined by the
optics of the unit. The I R thermometer in Figure 30-7 will correctly read the
temperature of Object A but will read a temperature lower than the temper-
ature of Object B if the background is of a lower temperature than the object.
 Body Temperature Monitors

Object A

Figure 30-7. Field of View of Optical Sensor.

TYMPANIC (EAR) THERMOMETERS

A tympanic thermometer measures body temperature by measuring the

IR energy emitting from the tympanic membrane in the ear. It is an inter-
mittent temperature measurement device. The tympanic membrane is a
good site to measure core temperature as it is vascularized by the carotid
arteries, which also perfuse the hypothalamus. Since tympanic thermometers
do not rely on thermal conduction, unlike most electronic thermometers,
they take less time to reach temperature equilibrium. A typical tympanic
thermometer takes less than 5 seconds to take a reading, and it is not subject
to external thermal disturbances (such as air temperature fluctuation).
Tympanic thermometers are recommended for temperature measurements
on unsettled patients.

BLOCK DIAGRAM OF A TYMPANIC THERMOMETER

A functional block diagram of a tympanic thermometer is shown in

Figure 30-8. In the diagram, the object is the tympanic membrane. The IR
energy radiated from the membrane is collected by the optical lens and fil-
ter. The detector converts the radiated energy to an electrical signal. This
electrical signal is processed to obtain the temperature. The following para-
graphs provide a more detailed functional description of the building blocks.
The optical path consists of a lens, a filter, and an optical light pipe. The
function of the lens and light pipe is to ensure that the FOV covers only the
tympanic membrane instead of other tissue such as the wall of the ear canal.
It also focuses the IR radiation to the detector. The filter allows only a cer-
 Biomedical Device Technology: Princ$les and Design

+Optical path >-

Electronic circuit

Object Light Lens F~lter

PIPe
Figure 30-8. Functional Block Diagram of IR Thermometer.

tain bandwidth of wavelength to reach the detector. The wavelength of the

I R spectrum lies between 0.7 and 20 pm. The bandwidth of a wideband IR
thermometer is several microns wide (e.g., 8-14 pm), whereas a narrowband
+
device allows only a single wavelength to reach the detector (e.g., 2.2 0.5
pm). Instead of using a single band wavelength, a "two color'' thermometer
uses two narrowbands to reduce errors due to nonunity emissitivity and I R
absorption in the optical path (e.g., due to water moisture in the atmosphere).
Figure 30-9 shows the pass band(s) of a wideband, narrowband, and "two
color" I R thermometer.
Several types of detectors can be used in I R thermometry. Common
detectors are pyroelectric sensors and thermopiles.
In a pyroelectric sensor, conductive material is deposited on the opposite
surfaces of a slice of a ferroelectric material. The ferroelectric material
absorbs radiation and converts it to heat. The resulting rise in temperature
changes the polarization of the material. The current flowing through the
external resistor connected across the two conductive surfaces is proportion-
al to the rate of change of temperature of the sensor. A shutter mechanism is
often installed to provide a controlled period of exposure to the radiation.
A thermopile is made up of a number of thermocouples connected in
series. Radiant energy landed on the hot junction area is first converted to
heat, creating a differential temperature between the hot and cold junctions
of the thermocouples. Each thermocouple generates a small voltage (on the
order of pV) according to this temperature difference. The output of the ther-
mopile is the summation of all the voltages from the thermocouples in the
sensor. An I R thermometer using a thermopile sensor requires another tem-
perature sensor, such as a thermistor, to measure the cold junction tempera-
ture. The construction of the thermopile in an I R sensor is shown in Figure
30-10.
To prevent sensing the ear canal temperature, some units use narrow
FOV optics with multiple scanning to detect the highest temperature in the
 Body Temperature Monitors

Figure 30-9. Detector Bandwidth of IR Thermometers.

ear canal. Within the ear canal, the hottest object is the tympanic membrane.
In most I R thermometers, the analog signal from the sensor is first digi-
tized and sent to the signal processor. The main function of the processor is
to determine the object temperature based on the signal from the sensor.
Calibration curves or lookup tables are stored in the unit's memory for this
purpose. Another function of the signal processor is to compensate for non-
ideal conditions such as E < 1.0 and provide offset to estimate the oral, rec-
tal, or core temperature from the tympanic reading.

POTENTIAL ERRORS I N TYMPANIC THERMOMETERS

Factors that affect accuracy unique to tympanic thermometers are:

If the FOV of the sensor includes other tissue inside the ear canal, it
will create a lower than normal temperature reading.
Nonstraight ear canal can prevent direct line of view of the detector to
the tympanic membrane. This situation often occurs in infants where
the ear canal is narrow and curved.
 Biomedical Device Technology: Princ$les and Design

Antimony , Bismuth

'k-

Heat sink for

Blackened area cold junctions
bonded to hot
junctions

Cold junctions
Figure 30-10. Construction of a Thermopile with Eight Thermocouples.

Too much hair or ear wax in the ear canal will block the tympanic
membrane from the detector.
Finally, some units provide software compensation factors to the tym-
panic temperature reading in order to estimate the temperature of dif-
ferent body temperature measurement sites (e.g., rectal, oral, etc.). In
these units, errors can be caused by incorrect selection of these com-
pensation factors.
 Chapter 3 1

PULSE OXIMETERS

State Lambert Beer's law.

Define SaOn, SpOz, SvOn, fractional oxygen saturation, and functional
oxygen saturation.
State the clinical applications of pulse oximetry.
Explain the principles of operation of pulse oximeters.
Describe the construction of a typical pulse oximeter sensor.
Sketch a typical block diagram of a pulse oximeter and explain the
functions of each block.
Discuss factors affecting signal quality and accuracy of pulse oximeters.
Differentiate between the clinical application of oxygen analyzers and
pulse oximeters.

CHAlTER CONTENTS

1. Introduction
2. Definition of Percentage Oxygen Saturation in Blood
3. Principles of Operation
4. Pulse Oximeter Sensor Probes
5. Functional Block Diagram
6. Errors in Pulse Oximetry
7. Differences Between Pulse Oximeters and Oxygen Analyzers
 Biomedical Device Technology: Princ$les and Design

INTRODUCTION

One of the main functions of the cardiopulmonary system is to deliver

oxygen and remove carbon dioxide to and from the cells. Hypoxia is a gen-
eral term describing the condition of lack of oxygen in the system. Acute hy-
poxia produces impaired judgment and motor incoordination. When hypox-
ia is long-standing, the symptoms consist of fatigue, drowsiness, inattentive-
ness, and delayed reaction time. More severe hypoxia can affect brain func-
tion and lead to death.
Until the early 1980s, blood oxygen saturation levels were measured by
drawing arterial blood samples from the patient and performing in vitro
analysis using laboratory co-oximeters (a multiwavelength spectrophotome-
ter). Pulse oximeters were developed in the early 1980s, providing a real-
time, continuous, and noninvasive means to monitor the changing level of
arterial blood oxygenation in patients; and allowing clinical intervention
before the occurrence of significant hypoxia. In 1986, pulse oximeters were
endorsed by the American Society of Anesthesiologists as a standard of care
for use whenever anesthesia is performed. Since then, pulse oximeters have
become the preferred method for measuring arterial oxygen saturation and
are used in most areas of hospitals.

DEFINITION OF PERCENTAGE OXYGEN

SATURATION IN BLOOD

The primary function of red blood cells is to transport oxygen from lungs
to tissue. This function is carried out by hemoglobin. When blood is circu-
lated into the lungs, oxygen is attached to hemoglobin, forming oxygenated
hemoglobin. Under normal conditions, hemoglobin in blood becomes al-
most fully saturated with oxygen before leaving the lungs. When blood is in
the capillaries, oxygen is released from the oxyhemoglobin and delivered to
the cells. The hemoglobin becomes deoxyhemoglobin. In studying oxygen
transport in blood, the terms %SaOz, %SPOY,and OIoSvOz are commonly
used. Here are their definitions:
Percent oxygen saturation of hemoglobin in arterial blood ("10Sa02)is
the percentage of hemoglobin in arterial blood that is bound with oxy-
gen; it is determined by analyzing an arterial blood sample with a co-
oximeter.
Percent oxygen saturation of hemoglobin in venous blood (OIoSvOz) is
the percentage of hemoglobin in venous blood that is bound with oxy-
gen; it is usually determined from a blood sample taken from or mea-
 Pulse Oximeters

sured in the pulmonary artery.

Percent oxygen saturation of hemoglobin in arterial blood (01oS~Oz)is
determined by a pulse oximeter (instead of from a blood sample by a
co-oximeter).
In addition to oxyhemoglobin and deoxyhemoglobin, there are two
other forms of hemoglobin. Carboxyhemoglobin is hemoglobin bound with
carbon monoxide. Methemoglobin is the oxidized form of hemoglobin.
Methemoglobin is incapable of binding with oxygen. A high percentage of
carboxyhemoglobin or methemoglobin compromises the oxygen-carrying
capacity of blood as there is less hemoglobin available to bind with oxygen.
Table 31-1 shows the normal values of 010SaOn and 01oS~Oawith the cor-
responding blood gas values of normal adults and neonates.

Table 31-1.
-
Wical Values of Blood Oxygen Level
-
-

Adult '%SaOn 96-98% Pa& 85-100 mmHg PaCOl 38-42 mmHg

0kS~Oz 70-75Yo P a s 35-40 mmHg P v C O ~ 41-51 mmHg
Neonates "koSa02 -94'Yn
-
-
Pa02 63-87 mmHg PaCOz 31-35 mmHg

PRINCIPLES OF OPERATION

The principle of pulse oximetry is based on Lambert Beer's law with dif-
ferential light absorption of two wavelengths. The wavelengths of the most
commonly used sources are red (660 nm) and the infrared (940 nm).
Lambert Beer's law states that for a substance of concentration C in a
fluid, the absorbance (A) of light due to the substance in the fluid, which is
defined as the natural logarithm of the ratio of incident light intensity (10) to
the transmitted intensity (I), is equal to the product of the absorptivity (a'),
the substance's concentration (C) in the fluid and the distance of the optical
path length (d). Figure 31-1 illustrates the concept.
For a mixture of two substances Xand Yin the fluid, the total absorbance
A is given by the sum of the absorbance due the substance X and the sub-
stance Yalone, or:

Two equations are used by pulse oximeter device manufacturers to cal-

culate oxygen saturation in blood. They are the fractional oxygen saturation
and the functional oxygen saturation equations.
(i) Fractional oxygen saturation (010OzHb) is equal to the ratio of the
 Biomedical Device Tachnology:Frinc@les and Desigrz

A=In-
lo =aC=aldC
I
A = absorbance
,I = intensity of incident beam
I = intensity of transmitted beam
C = concentration of substance
a = a'd
a' = absorpbvity (a constant for the substance)
d =thickness of substance (optical path length)
Figure 31-1. Lambert Beer's Law.

concentration of oxyhemoglobin in blood to the sum of concentrations of all

types of hemoglobin in blood.

where:
C O ~=Hthe~ concentration of oxygenated hemoglobin in arterial blood,
C H H=~the concentration of deoxygenated hemoglobin in arterial blood,
CCOH~ = the concentration of carboxyhemoglobin in arterial blood, and
Cmcl~b= the concentration of methemoglobin in arterial blood.
(ii) Functional oxygen saturation (010SaO2)is equal to the ratio of the
concentration of oxyhemoglobin in blood to the sum of the functional hemo-
globin concentrations. That is, the concentrations of the oxyhemoglobin and
the deoxyhemoglobin, which are responsible for the oxygen transport.

It is obvious from the preceding equations that the functional value is

higher than the fractional value for the same blood sample. For most
patients, the concentration of carboxyhemoglobin as well as that of the
methemoglobin is negligible. In a healthy individual, the difference between
0/oSaO:! and 01002Hbis less than 3%.
Figure 31-2 shows the absorption spectrum of oxyhemoglobin (OnHb)
and deoxyhemoglobin (HHb).To measure the 010SaO2(or OIoOsHb) in a blood
sample, two light sources of wavelengths A1 and A2 are used. In the blood
sample, let:
Co be the concentration of oxyhemoglobin (OzHb) in blood, and
 Pulse Oxirneters

O2Hb

HHb

1 b
600 800 I,000 Wavelength
Figure 31-2. Absorption Characteristics of Oxy- and Deoxyhemoglobin.

Cd be the concentration of of deoxyhemoglobin (HHb).

At wavelength A 1, using Lambert Beer's law:

where A1 = the total absorption due to wavelength A1

AIO= the absorption of oxyhemoglobin due to wavelength A1
A I =~ the absorption of deoxyhemoglobin due to wavelength A1
aio = the product of the optical path length and the absorptivity of oxyhe-
moglobin due to wavelength A1,and
aid = the product of the optical path length and the absorptivity of deoxyhe-
moglobin due to wavelength A1
At wavelength A2, using Larnbert Beer's law:

One can solve equations (1) and (2) for Co and Cd if Ai, A2, ale, aid, a20,
a dare known. Knowing Co and Cd, the oxygen saturation can be computed.
Using the functional oxygen saturation equation (Equation 2):
 Biomedical Device Technology: Principles and Design

In practice, the light beams travel through the tissue and are absorbed
not only by the hemoglobin but also by other tissues (such as bone, muscle)
in the light path. In addition, as the diameters of the capillaries are pulsating
according to the blood pressure, the optical path length is not a constant.
Therefore, aio, aid, mo, and a24 which are the product of the absorptivity and
optical path length are not exactly constant values and hence Co and Cd can-
not be computed analytically.
Figure 31-3 shows the absorption waveform measured by the light sen-
sor in a pulse oximeter probe. A red beam (X1 = 660 nm) and an infrared
(A2 = 940 nm) beam are commonly used. The solid and dotted waveforms
are results of the absorption characteristics of each of the beams.

Waveform due to artertal Red and

pulsatile s~gnal(A1= 660 nm) IR LED
Waveform due to arterial

sensor

b
Time
Figure 31-3. Absorption Signal from Pulse Oximeter Finger Probe.

Most pulse oximeter manufacturers derive the 010SaOa values from the
optical intensity ratio (r) of the transmitted intensity of the red (Ird) and
infrared beam (h)measured by the sensors in the probe.
r = - 1rd
Iir '

In most cases, an empirical equation or a lookup table between rand the

%SaOz values is established so the OIoSaOn value can be determined from the
measured I&and &. This correlation between rand the %SaOz is verified sta-
tistically by simultaneously reading the pulse oximeter output and drawing
an arterial blood sample from the patient and analyzing the sample by a co-
oximeter. Pulse oximeters are calibrated using arterial blood samples (blood
oximeter) using either the OIoOzHb (fractional) or 010SaOa (functional) equa-
tions.
 Pulse Oximeters

PULSE OXIMETER SENSOR PROBES

Many different types of sensor probes are used in pulse oximetry. A typ-
ical probe consists of two light-emitting diodes (LEDs),one emitting red light
and the other emitting infrared. These LEDs are pulsed alternately to send a
beam of light through the underlying tissues (see the top right figure in Figure
31-3). A photodetector in the probe on the other side of the tissue picks up
the transmitted light signal and sends it to the processing circuits. Probes can
be classified as reflectance or transmittance, disposable or reusable, or by
their sensing locations. A disposable probe is one that will be discarded after
being used on a single patient (however, some sites reuse some probes that
are labeled "single use" for cost saving). The LEDs and the photodetector are
mounted on each end of a flexible strip. The strip is applied on and often
taped over the tissue (Figure 31-4). A reusable probe usually has a more
robust and rigid cover to protect the LEDs and the detector. A transmitting
probe has the LEDs on one side and the detector on the other side of the
capillary bed. The light is transmitted through the capillary bed and tissues.
On the other hand, in a reflecting probe, the detector is placed on the same
side as the LEDs. As the light penetrates the tissue, some is absorbed and
some is reflected back to the surface. The detector picks up the reflected sig-
nal and sends it to the processor. Pulse oximeter probes in theory can be
placed over any part of the body with capillaries. However, common sites for
transmitting probes are the index finger and thc earlobe. For infants, probes
are often taped to the big toe. Transmitting probes are usually placed on the
forehead of the patient.

FUNCTIONAL BLOCK DIAGRAM

Figure 31-5 shows a typical functional block diagram of a pulse oxime-

ter. It consists of:
A probe consisting of a red LED, an infrared LED, and a photodetec-
tor.
A timing control circuit to sequence the LEDs and synchronize them
with the photodetector. There are three phases in one timing cycle:
Red "on* and I R "off," Red "off" and I R "on," and both LEDs "off."
The latter is to measure the dark signal to eliminate the effects of the
ambient light.
Analog and digital electronics to amplify and process the signal.
A processor to compute the transmitted red and infrared light intensi-
ty ratio and match the 010Sa.02 from the lookup table. It can also derive
 Biomedical Device Technology: Princ$les and Design

/IR LED Detector

1
Cable to oxirneter
(Disposable Probe)

connector

Red and Skin

Red and IR LED Detector surface
Capillaries IR LED

-
Detector capillaries

(Transmittance probe) (Reflectance Probe)

Figure 31-4. Pulse Oximeter Probes.

the heart rate from the pulsating waveform and compares the mea-
sured values (heart rate, 010Sa0~)
to the alarm settings.
A display to show the 010SaOn values, the alarm limits, and the heart
rate. A plethysmograph showing the detected signal strength is often
displayed to provide the user an idea of the signal to noise information
of the measurement. A strong signal level indicates that the measured
value is reliable. A plethysmograph can be a waveform similar to the
absorption waveform shown in Figure 31-3 or simply a one-column
bar graph proportional to the detected signal strength.

ERRORS IN PULSE OXIMETRY

Although we can empirically establish an accurate relationship under

ideal situations between the optical intensity ratio (r) and 010Sa02, in the pres-
ence of patient motion or other interference, the optical densities will
inevitably include noise components (NJ. Therefore, in the presence of noise,
+
the measured beam intensity I = S N where S is the desired signal and N
is the noise. The optical intensity ratio (equation 5) is now rewritten as
 Pulse Oximeters

+
n Timing
control

+
Analog to
Calibration

Probe ---+ Amplifier + digital +Processor + Display

converter

Figure 31-5. Block diagram of Pulse Oximeter.

r = Srd + Nrd
Sir + Nir
A well-observed source of noise is the change in light absorption caused
by patient motion. The movement changes the optical path length of blood
vessels and tissues. In a poor signal to noise ratio situation, the noise level
becomes significant in the optical intensity ratio. This will increase the error
in the derived 010Sa02.If the noise (N) component is much larger than the sig-
nal(4, i.e. N >> S, the optical intensity ratio in equation (6) then becomes
r = - Nr d
Nir *

Under such conditions, if the noise level in the red and IR regions are
similar (i.e., Nrd =" Ah), the optical intensity ratio r will approach unity.
For most systems with a good signal to noise ratio, r = 1.0 corresponds to
a 010SaOn of about 82%. For that reason, a pulse oximeter working under
noisy conditions will tend to report a lower oxygen saturation reading. To
maximize the signal to noise ratio, the transmitted beam intensity should be
measured during the systolic region of the blood pressure cycle. There have
been some reported successes by manufacturers using special digital signal
processing techniques such as adaptive filtering or signal extraction to mini-
mize the effect of noise in pulse oximetry.
In summary, errors in pulse oximetry measurement are due to the fol-
,'
- lowing causes:
Poor perfusion-A patient suffering from poor perfusion usually has
lower than normal blood pressure. A lack of blood in the capillaries
will decrease the signal to noise ratio and therefore increase the error
of the measurement.
Excessive signal attenuation-Patients with dark skin pigment or too
thick tissue (e.g., skin) at the measurement site will decrease the signal
penetration (decrease the detector signal level) and increase measure-
 Biomedical Device Technolopy: Princ$les and Design

ment error.
External interference-EM1 and ambient light can introduce errors in
measurement. There were reported incidents that flashing light and
fluorescent light sources were misinterpreted by machines as pulsating
red or IR signals. To avoid external light interference, pulse oximeter
probes are usually designed with a cover to block external light from
reaching the sensor.
Motion-Motion will cause changes in the optical path length, which
will produce measurement errors.
Substances in blood-Some substances in the bloodstream may affect
the absorption of the light sources. A high level of dyshemoglobin in
carbon monoxide poisoning, low hematocrit counts of an anemic
patient, and artificial dyes in a patient's blood can all affect the accu-
racy of the measurement.

DIFFERENCES BETWEEN PULSE OXIMETERS

AND OXYGEN ANALYZERS

Oxygen analyzers and pulse oximeters are the two devices commonly
used today to monitor a patient's oxygen level in the clinical environment.
An oxygen analyzer measures the percentage of oxygen gas in a gas mixture
such as the inspired air of a patient. Oxygen analyzers are usually attached
to the patient breathing circuit. A pulse oximeter measures the oxygen con-
tent in the patient's blood. Measuring oxygen saturation level in blood can
detect hypoxia even before other signs such as cyanosis or hyperventilation
are observed. Table 31-2 summarizes the main differences between the two
devices.

Table 31-2.
Oxygen Analyzer and Pulse Oximeter Comparison.
--- -
-

Oxygen Anabzr - - Pulse Oximeter

Principle of operation Electrochemical transducer Lambert Beer's law
Parameter sensed Partial pressure of 0 2 in airway Oxygen saturation in blood
Hypoxia detection Detects oxygen deficiency Detects insufficient oxygen
in inhaled air in blood stream

Both devices are often used together to detect insufficient oxygen to the
patient. They provide complementary protection against hypoxia. For exam-
 Pulse Oximeters

ple, in anesthesia, an oxygen analyzer is connected to the inspiratory limb of

the patient breathing circuit to sound an alarm on low oxygen level in the
patient's inspired gas. A pulse oximeter is connected to the patient (e.g.,
using a finger probe) to detect the actual level of oxygen in the patient's
bloodstream.
 Chapter 32

END-TIDAL CARBON DIOXIDE MONITORS

State the clinical applications of end-tidal COa monitors.

Sketch and explain the change in COz concentration in expired air.
Explain the principles of operation of end-tidal COa monitors.
Differentiate between mainstream and sidestream end-tidal COXmoni-
toring.
Identify the functional components and the construction of a typical
mainstream end-tidal Con sensor.
Sketch the functional block diagram of a typical sidestream COa moni-
tor.
Discuss factors affecting the signal quality and accuracy of end-tidal
COz monitors.

CHAPTER CONTENTS

1. Introduction
2. Carbon Dioxide Concentration Waveform (Capnogram)
3. Principles of Operation
4. Mainstream Versus Sidestream Monitoring
5. Errors in Capnography
 End-Tidal Carbon Dioxide Monitors

INTRODUCTION

Carbon dioxide is a by-product of cellular metabolism and is removed

from the body through the circulatory and respiratory systems. The concen-
tration of COz in the exhaled air reflects the metabolic rate and indicates the
status of the pulmonary and circulatory systems. Carbon dioxide is continu-
ously produced in the body and transported to the alveoli. The two factors
that determine the alveolar concentration of COz are the rate of transport of
COYfrom the blood to the alveoli and the rate of removal of COz from the
alveoli by alveolar ventilation. The alveolar COz concentration is directly
proportional to the rate of Con excretion and inversely proportional to the
alveolar ventilation.
A COYmonitor can noninvasively measure the concentration of C02 in
breathing air. It is primarily used in operating rooms to monitor patients'
ventilation under general anesthesia. It can detect breathing circuit discon-
nection, airway leaks, and improper placement of an endotracheal tube. In
intensive care units, it can be used together with other physiological moni-
toring parameters to evaluate a patient's cardiopulmonary function.

CARBON DIOXIDE CONCENTRATION WAVEFORM

(CAPNOGRAM)

The expired air in a breathing cycle is a combination of the dead space

air and alveolar air. Figure 32-1 shows the changes of COz concentration in
the expired air during the course of breathing. The expired air at the begin-
ning of exhalation is dead space air, which is inspired air saturated with mois-
ture. As expiration goes on, more and more alveolar air becomes mixed with
the dead space air until the dead space air has been totally washed out. At
the end of expiration, the expired air contains 10O0Ioalveolar air. For a nor-
mal adult breathing in atmospheric air, the partial pressure (or concentra-
tion) of the CO2 in the inspired air (atmospheric air) is 0.3 mmHg (or 0.04%)
and that of the alveolar air is about 40 mmHg (or 5.3%). At the beginning of
expiration, the COz concentration is 0.04°/o and it rapidly rises until reaching
a plateau. At the end of the expiration, the C02 concentration is at its max-
imum (about 5%).
A normal capnogram includes a return to zero (0.04%) baseline, a sharp
upstroke, and a relatively horizontal alveolar plateau. Significant deviations
from this morphology suggest an abnormality in the patient or the gas deliv-
ery equipment.
 Biomedical Deuice Technolog: Princ$les and Design

+Inspiration + Expiration 3- ' 0

Time (s)

Figure 32-1. Con Concentration in Breathing Air.

PRINCIPLES OF OPERATION

The principle of operation of an end-tidal COz monitor is similar to that

of a pulse oximeter. COz monitors use infrared spectroscopy to measure the
concentration of COz in the exhaled gas. In normal exhaled gas, only COa
and water vapor contribute to the absorption of I R radiation. However, dur-
ing anesthesia procedures, anesthetic gases such as N20 may affect the IR
absorption. Figure 32-2 shows the IR absorption spectrum of water vapor,
carbon dioxide, and nitrous oxide. Based on Lambert Beer's law, the con-
centration of COa in the gas is proportional to its absorption. Selecting the
proper wavelength and correcting the effects of other absorptions allows the
monitor to accurately determine the COa concentration.
A typical setup to measure COa concentration is shown in Figure 32-3.
Infrared radiation passing through the gas mixture is absorbed by the COa
in the exhaled gas sample. The I R beam then passes through two filters; the
first filter is within the I R frequency spectrum of Con absorption and the sec-
ond is outside. Behind each filter is an I R detector. The detector behind the
first filter is for the data channel and the detector behind the second filter is
for reference. The reference signal is to compensate for signal fluctuations
such as variation of intensity of the light source.

MAINSTREAM VERSUS SIDESTREAM MONITORING

There are two types of capnographs, mainstream and sidestream, which

are differentiated by the location of their COa sensors. A mainstream capno-
graph has the sensor located right in the patient breathing circuit. Figure
 End-Tidal Carbon Dioxide Monitors

4 Wavelength (pm) 10
Figure 32-2. IR Absorption Spectra of HzO, Nz0 and COY.

Figure 32-3. Functional Components of COz Sensor.

32-4 shows a COYsensor of a mainstream capnograph connected to the

patient breathing circuit using a special airway adaptor. The windows on
both sides of the airway adaptor allow I R from the light source to reach the
detectors. As COz gas mixture passes through the airway adaptor, the IR
radiation of the pulsing source (e.g., 48 Hz) is partially absorbed by the COz
in the gas mixture. It then passes through the two optical filters before reach-
ing the photodetectors. One filter is at the middle of the IR absorption spec-
trum (e.g., 4.2 pm) for carbon dioxide and the other is outside (e.g., 3.7 pm).
To accurately measure the level of COz in the gas mixture, the outputs of the
detectors are sampled simultaneously and the level of COz is determined
from the ratio of the data and reference channels. The ratio is then compared
to a lookup table in memory to determine the concentration of C02.
As the warm expired air is saturated with water vapor, water will con-
dense on the cooler windows of the sensor, causing errors in the IR absorp-
 Biomedical Device Technology: Priac$les and Design

tion measurement. To prevent condensation from forming on the window, a

heater and a thermistor inside the sensor maintain a sensor temperature
above the body temperature (eg., 40°C).

Filters Wlndows Airway adaptor

IR detectors , . /
Light source
/

Heater

Cable

Figure 32-4. Internal View of a Mainstream COz Sensor.

The sensor of a sidestream capnograph is inside the machine. The prin-

ciple of C02 detection is similar to its mainstream counterpart. A small-
diameter air sampling line connects the patient breathing circuit to the sen-
sor inside the machine. During measurement, breathing gas sample is drawn
into the sensor from the breathing circuit by an air pump. To prevent con-
densation, a water trap is located at the end of the air sampling line at the
entrance of the machine. A heater is also used to raise the temperature of the
sensor chamber to prevent water condensation. For both mainstream and
sidestream monitors, the derived COz concentration (or partial pressure)
must be temperature and pressure corrected as it is measured at temperature
and pressure different from those inside the patient's body. Figure 32-5
shows a functional block diagram of a sidestream Con monitor.

ERRORS IN CAPNOGRAPHY

As discussed previously, moisture as well as its condensation absorb IR

radiation. Water condensation is a major problem for sidestream monitors as
it can block the small-bore air sampling tubing, causing a pressure and flow
reduction in the circuit. Since the partial pressure of Con is pressure-depen-
dent, any pressure fluctuations will introduce errors in the measurements.
 End-Tidal Carbon Dioxide Monitors

IR source and
reflector

Figure 32-5. Sidestream COz Monitor.

When a COz monitor is used during anesthesia procedures, halogenated

gases and nitrous oxide are present in the patient's exhaled gas. These gases
absorb I R radiation and may cause measurement errors if they are not prop-
erly accounted for.
Although the presence of oxygen does not affect IR absorption, the col-
lision of Oz and C02 molecules will broaden the absorption peak and create
a falsely low COz concentration reading.
Capnographs may not be able to accurately measure the alveolar COz
concentration of neonates or patients with shallow breathing due to the rel-
ative large dead space volume. In addition, the gas sampling flow rate
required in sidestream monitors may be too high (e.g., as high as the minute
volume) for neonates.
 Chapter 33

ANESTHESIA MACHINES

OBJECTIVES

State the functions of an anesthesia machine.

Name and explain the functional building blocks of an anesthesia
machine.
Sketch the gas circuit/piping diagram of a continuous-flow rebreathing
anesthesia machine.
Trace the oxygen, nitrous oxide, and anesthetic gas flow in a typical
anesthesia machine and its patient breathing circuit.
List potential hazards associated with the use of an anesthesia machine
and the built-in features in anesthesia systems to minimize such risks.

CHAPTER CONTENTS

1. Introduction
2. Principles of Operation
3. Gas Supply and Control Subsystem
4. Breathing and Ventilation Subsystem
5. Scavenging Subsystem
6. Major Causes of Injury and Preventive Measures

INTRODUCTION

Since Dr. Crawford Long of Jefferson, Georgia, administered the first

ether anesthetic in 1842 for the painless removal of a neck tumor, anesthesi-
486
 Anesthesia Machines

ology has evolved into a branch of medicine that is concerned with the
administration of medication or anesthetic agents to relieve pain and support
physiological functions during a surgical procedure. In a surgical procedure,
an analgesic or anesthetic agent is administered to the patient for pain relief.
For a major or long surgical procedure, the patient may undergo general
anesthesia. General anesthesia is a reversible state of unconsciousness pro-
duced by anesthetic agents in which motor, sensory, mental, and reflex func-
tions are lost. While depressing the cerebral cortex to achieve anesthesia, the
anesthetic agents may cause respiratory depression or even respiratory
arrest. An anesthesia machine is a collection of medical instrumentations to
assist the anesthetist to induce and maintain anesthesia. It serves three major
functions:
1. Dispense a controlled mixture of gases consisting of anesthetic agents
(such as halothane, enflurane, nitrous oxide, etc.) and oxygen to
anesthetize the patient during surgery.
2. Assist patient's respiration (ventilation) when normal breathing is
compromised due to the anesthetic effect.
3. Monitor the patient's condition, such as vital signs and depth of anes-
thesia, during the surgical procedure.
Physiological parameter monitors with alarm functions are often inte-
grated into the anesthesia machine. Automatic record keeping of vital signs
and ventilation parameters, networking capabilities, and data management
functions are available in modern anesthesia systems.
This chapter introduces the backbone of a typical anesthesia machine,
which includes the gas supply and control, breathing and ventilation, and
scavenging subsystems. Technology and instrumentation in physiological
monitoring is discussed in other parts of this book.

PRINCIPLES OF OPERATION

A basic anesthesia machine consists of three main subsystems:

1. Gas supply and control
2. Breathing and ventilation
3. Scavenging
Patient monitoring equipment may be considered as the fourth subsys-
tem (which is not discussed in this chapter). The most commonly used anes-
thesia machine is the continuous-flow rebreathing anesthesia machine. In
this type of machine, the exhaled gas from the patient is supplied back to the
patient after it is processed and mixed with a proportion of fresh anesthetic
 Biomedical Device Technology: Princ$les and Design

gas. Figure 33-1 shows a simple functional block diagram of such a machine.
The gas supply and control block takes oxygen and nitrous oxide from the
wall outlet and combines them with an anesthetic agent to produce a mix-
ture of anesthetic gas. This gas mixture, being regulated to an appropriate
level of flow and pressure, is supplied to the breathing and ventilation circuit.
The function of the breathing and ventilation circuit is to deliver the gas mix-
ture to the patient as well as to remove the expired gas from the patient. The
scavenging block is to prevent the exhaled or released anesthetic gas from
polluting the operating room environment. The scavenging system captures
the waste anesthetic gas and discharges it safely outside the operating room.

02.N20, and Gas supply Breathing and

anesthetic agent and control ventilation b To patient

Scavenging

Figure 33-1. Functional Block Diagram of a Basic Anesthesia Machine.

Figure 33-2 shows a basic anesthesia machine. The following sections

describe the principles of operation of each of the three subsystems.

GAS SUPPLY AND CONTROL SUBSYSTEM

Figure 33-3 illustrates a gas piping diagram showing the major compo-
nents of the gas supply and control subsystem of an anesthesia machine.
Under normal operation, oxygen and nitrous oxide gases are supplied
from the wall piped gas outlets to the oxygen and nitrous oxide gas inlets of
the machine. The pressure of the gases at the wall outlets is about 50 psi.
Oxygen flow through a check valve (one-way valve) is reduced to about 16
psi by the second-stage oxygen regulator before it reaches the flow control
valve of the oxygen flowmeter. Nitrous oxide gas from the wall outlet passes
through the pressure-sensing shutoff valve and reaches the nitrous oxide flow
control valve of the nitrous oxide flowmeter. The shutoff valve is held open
by the oxygen pressure, which is normally at 50 psi. If the oxygen pressure
drops to below 25 psi, the valve will shut off the nitrous oxide supply to the
 Anesthesia Machines

Flowmeter Vaporizors

linder
yoke

Figure 33-2. Basic Anesthesia Machine.

machine. In addition, the shutoff valve will discharge the oxygen stored in
the alarm cylinder through a whistle to alert the anesthetist to the failing oxy-
gen supply. This shutoff and alarm mechanism will protect the patient from
unknowingly breathing in a low oxygen level gas mixture in case of supply
oxygen failure.
By adjusting the flow control valves, the anesthetist can achieve a suitable
mix and flow of oxygen and nitrous oxide gas mixture. The gas mixture then
flows through a calibrated vaporizer, where it picks up a selected amount of
an anesthetic agent before it is delivered to the patient via the breathing and
ventilation circuit. An oxygen flush valve is available to flush the patient cir-
cuit during setup.
Under rare circumstances in which piped-in wall gases are not available,
oxygen and nitrous oxide supplies are automatically switched to gas cylin-
ders mounted on the machine. There are usually two cylinders of oxygen
and two cylinders of nitrous oxide mounted on the hanger yokes on the sides
or the rear of the anesthesia machine. One cylinder of each gas is on stand-
by (turned on) and the other one serves as backup in case the first cylinder
is depleted. As the maximum pressure of oxygen and nitrous oxide from
 Biomedical Device Technology: IPinc$les and Design

Machine gas outlet

to breath~ngcircuit
-
4 -+

2nd-stage oxygen

cyi~nderhanger

-
yokes

ntuous oxide p s l~ur oxygen gas lmne v

- mtrcd imestheuc gas lme

Figure 33-3. Major Components of the Gas Supply and Control Subsystem.

their cylinders are 2,200 psi and 750 psi, respectively, cylinder pressure reg-
ulators are installed to reduce the pressure of these gases to about 50 psi.
Pressure gauges at hanger yokes provide an indication of whether the cylin-
ders are full or empty. As oxygen in the gas cylinder is in gaseous state, the
level of gas in the cylinder is proportional to the pressure in the cylinder.
However, it is not possible to tell the amount of nitrous oxide left in the cylin-
der by reading the cylinder pressure. Since nitrous oxide is in liquid form
inside the cylinder, the cylinder pressure will start to drop only when it is
almost empty (no more liquid N20 in the cylinder).
A vaporizer is used in the gas supply and control circuit to introduce a
selected concentration of the anesthetic agent into the oxygen and nitrous
oxide mixture to form the anesthetic gas. Two types of vaporizers are found
on anesthesia machines: the conventional variable-bypass vaporizers and the
electronically controlled vaporizers. Variable-bypass vaporizers are used to
deliver liquid anesthetic agents such as halothane, enflurane, isoflurane, and
servoflurane. Anesthetic agents (such as desflurane) that are in gaseous form
under room temperature require the electronically controlled vaporizer.
There is usually more than one vaporizer (with a different agent in each)
mounted on the output manifold of the anesthesia machine. However, only
one vaporizer can be active at one time.
Figure 33-4a shows the simplified construction of a variable-bypass
vaporizer. The oxygen and nitrous oxide gas mixture enters the vaporizer
 Anesthesia Machines

from the inlet. It is then split into two flow paths, one into the vaporizing
chamber and the other through a bypass into a mixing chamber. The per-
centage of the total flow into the vaporizing chamber is determined by the
position of the agent concentration control valve. The gas mixture flowing
into the vaporizing chamber flows over a reservoir of liquid anesthetic agent,
picks up the vaporized agent, and exits the vaporizing chamber. The gas then
meets and mixes with the bypassed gas and flows to the vaporizer outlet. The
concentration of anesthetic agent in the final gas mixture is higher when a
larger volume of gas is allowed to flow into the vaporizing chamber.
Temperature-sensing flow control devices are necessary to compensate for
temperature variations during the procedure.
In an electronically controlled vaporizer, the anesthetic agent is pressur-
ized into its liquid state and heated inside the agent chamber. The pressur-
ized vapor of the anesthetic agent is released through a regulating valve into
the mixing chamber. The concentration of the agent in the gas mixture is
adjusted by the control valve. To maintain a constant agent concentration,
the electronic controller of the regulating valve determines the valve's posi-
tion from a number of factors: the selected agent concentration, the gas flow
rate, the pressure and temperature inside the vaporizer chamber, as well as
the pressure and temperature of the inlet gas mixture.

Agent concentrat~on
control Adjustable pressure rel~efvalve
(agent concentration control)
Vapor~zeroutlet

(a) ~ i ~anesthetic
i d agent (b)
Figure 33-4. (a) Variable-Bypass Vaporizer, (b) Electronic Vaporizer.
 Biomedical Device Technology: Principles and Design

BREATHING AND VENTILATION SUBSYSTEM

The function of the breathing and ventilation subsystem of an anesthesia

machine is to deliver the anesthetic gas mixture (which is a mixture of oxy-
gen, nitrous oxide, and an anesthetic agent) to the patient. During a proce-
dure when the patient's respiration function is suppressed, a mechanical ven-
tilator in controlled ventilation mode is used to ventilate the patient. Most
anesthesia machines deliver a continuous flow of anesthetic gas and oxygen
mixture to the patient. Two types of breathing circuits are used: the circle sys-
tem and the T-piece system. Both types allow a certain level of rebreathing
to conserve moisture and heat as well as anesthetic agents.
Figure 33-5 shows the circle breathinghentilation subsystem of an anes-
thesia machine under ventilator mode. The ventilator valve is used to select
between manual breathing mode using the breathing (or reservoir) bag and
automatic ventilation mode using the built-in mechanical ventilator. The

Fresh gas CO2 APL valve Pressure-limiting

Ventilation valve
supply inlet absorber valve
I \

tube
I
'----------------I

Figure 33-5. Patient BreathingNentilation and Scavenging Systems.

 Anesthesia Machines

scavenging subsystem for waste anesthetic gas removal is also shown in the
diagram.
Figure 33-6 shows the machine in manual breathing mode. The Y-con-
nection of the patient breathing circuit is connected to a face mask covering
the patient's mouth and nose or to an endotracheal tube inserted into the tra-
chea of the patient. During patient inhalation, fresh gas from the anesthesia
machine enters the inspiratory limb of the breathing circuit into the lungs of
the patient (flow direction in solid arrows). During exhalation (flow direction
in dotted arrows), expired gas from the lungs goes through the expiratory
limb of the breathing circuit into the breathing bag. A pair of check valves is
used to prevent reverse gas flow in the inspiratory and expiratory limbs of
the breathing circuit. These check valves are also referred to as pop-off
valves due to their construction (they consist of a circular disk sitting on top
of a circular opening of the breathing circuit). When positive pressure is cre-
ated in the circuit by manually squeezing the breathing bag, the gas collect-
ed in the bag is driven through a COz absorption canister, through the inspi-
ratory limb of the breathing circuit, and back to the patient. The canister con-
tains a COz absorber (such as soda lime) to remove COz from the rebreathed

-+ Inhalation gas flow

----- +, ~xhalacongas flow

Figure 33-6. Patient Circuit Under Manual Breathing Mode.

 Biomedical Device Technology: Principles and Design

--+Inhalation gas flow

----- + Exhalation gas flow

Figure 33-7. Patient Circuit under Ventilator Mode.

gas. The maximum pressure in the breathing circuit is limited by the

adjustable pressure-limiting (APL) valve located near the patient bag. The
APL valve is adjusted by the anesthetist during the procedure to maintain a
slightly inflated breathing bag.
When the machine is in ventilator mode (Figure 33-7), expired gas flows
into the bellow of the ventilator instead of into the patient breathing bag.
During inhalation, the positive pressure of the control gas compresses thk
bellow in the ventilator, forcing the accumulated gas in the bellow to flow
through the COz absorber, and then through the inspiratory limb of the
breathing circuit to the patient. The maximum pressure in the breathing cir-
cuit is limited by the pressure-limiting valve of the ventilator. In the con-
trolled mode, the ventilator is cycled to deliver a fixed volume of gas to the
patient at a fixed time interval. The volume and frequency of ventilation are
set by the anesthetist.
No COz absorber is used in the T-piece design (Figure 33-8). Fresh gas
is mixed with rebreathed gas before entering the patient's lungs. The per-
centage of rebreathing is controlled by the flow rate of fresh gas from the gas
supply and control system. The exhaled anesthetic mixture leaves the circuit
 Anesthesia Machines

Fresh gas
supply
To patient

Figure 33-8. Rebreathing Circuit Using T-Piece Design.

through an APL valve. The rate of elimination of exhaled COz in the

rebreathed air mixture is proportional to the flow rate of the fresh gas. This
design is often used in pediatric anesthesia.

SCAVENGING SUBSYSTEM

Prolonged personal exposure to anesthetic agents, including nitrous

oxide, has been shown to be related to some illnesses such as liver disease
and premature infant birth. Exposure to anesthetic agents is classified as an
occupational hazard for operating room personnel. The scavenging subsys-
tem is designed to capture and remove waste anesthetic gases from the oper-
ating room to reduce such health hazards. Instead of being released into the
operating room, waste anesthetic gases are collected from the exhausts of the
APL valve and pressure-limiting valve of the breathinghentilation circuit
(Figure 33-9). Scavenging systems remove gas by connecting to either a vac-
uum or a passive exhaust. A vacuum scavenging system uses the wall suction
in the operating room, whereas the passive exhaust scavenging system con-
nects to the exhaust of the room air ventilation system. System pressure
adjustment is critical to ensure that the scavenging system is of just enough
negative pressure to prevent waste anesthetic gases from being released into
the operating room but not so much as to remove too much of the patient
breathing gases. A scavenging bag is attached to act as a reservoir to absorb
fluctuations. As a safety measure, a pair of pressure-limiting valves are placed
near the scavenging bag to limit the pressure to within f0.5 cmHzO of the
atmospheric pressure. When the gas flow and APL valve are properly adjust-
ed, no anesthetic gas should be released to the atmosphere from these valves.
 Biomedical Device Technology: Princ$les and Design

or room exhaust

Scavenging gas flow

Figure 33-9. Scavenging Subsystem.

MAJOR CAUSES OF INJURY AND PREVENTIVE MEASURES

A patient under general anesthesia is unconscious and cannot react or

signal for help. Patients under general anesthesia are vulnerable to operating
errors or malfunctions of instrumentation. A list of potential safety hazards to
patients under general anesthesia follows.
Insufficient oxygen supply to patient
Insufficient carbon dioxide removal from the rebreathing gas
Excessive or wrong anesthetic agents being delivered to the patient
Excessive breathing circuit pressure leading to trauma to the lung
Introduction of foreign particles into the airway
Insufficient oxygen supply to the patient will lead to hypoxia. An oxygen
analyzer is used to monitor the oxygen level of the inspired gas. To detect
hypoxia in the patient, a pulse oximeter probe is connected to the patient to
measure the oxygen saturation level in the arterial blood. An oxygen ratio
monitor on the flow regulators of the anesthesia machine prevents the oxy-
gen level in the gas mixture from being set below 30%. Touch-coded oxygen
 Anesthesia Machines

and nitrous oxide control knobs prevent the anesthetist from mistaking the
nitrous oxide control for the oxygen control. Backup cylinders of oxygen
and nitrous oxide are in place to ensure an uninterrupted supply of oxygen
and nitrous oxide during a procedure. Color-coded gas cylinders and hoses
(oxygen-white in Canada or green in the U.S., nitrous oxide-bIue) are used
to prevent reversed connection of the gases. Different sizes of hose are used
to prevent misconnection of gas lines (breathing circuit-22 mm, fresh gas
supply and patient Y-15 mm, scavenging hose-19 mm). A diameter-indexed
safety system (DISS) using different diameters of connectors prevents an
oxygen cylinder from being connected to nitrous oxide pipelines. A pin-
indexed safety system prevents the wrong gas cylinder from being connect-
ed to a cylinder yoke of another gas. Vital sign monitors are used to assess
the patient's physiological condition.
Failure to remove carbon dioxide from the patient can lead to hyper-
capnia. An end-tidal COs monitor can detect an abnormal level of carbon
dioxide in the patient's exhaled gas. A low oxygen saturation level in arteri-
al blood (by a pulse oximeter) may also be an indication of excessive carbon
dioxide in the patient's system. Carbon dioxide absorbers are used to
remove CO2 from the exhaled gas.
Delivery of a wrong anesthetic agent can be fatal. The filling spout of a
vaporizer is keyed to accept only the bottle of the correct agent. Vaporizers
on an anesthesia machine are interlocked to allow only one vaporizer to be
turned on at one time. To ensure a correct mixture of anesthetic gases being
supplied to the patient, agent monitors are built into anesthesia machines to
monitor the concentration of the anesthetic agent during the procedure.
Bispectral (BIS) index monitoring, a special EEG measurement, may be
used to assess the depth of anesthesia (level of consciousness) of the patient.
Overpressure in the airway will create injury to the patient. Pressure
monitors and overpressure alarms are installed on all machines. The APL
valve and pressure-limiting valves limit the maximum pressure in the patient
air circuit. Filters and traps are used to prevent patient injury from foreign
particles.
Table 33- 1 summarizes the preceding discussion. Daily functional verifi-
cations of anesthesia machines as well as their accessories are performed by
the operating room staff, whereas performance inspection is done by bio-
medical engineering personnel during scheduled inspections.
 Biomedical Device Technology: Princ$les and Design

Table 33-1.
--
Summary of Hazards and Mitigation Related to Anesthesia Machines
- - --

Potential Hamrds Methods to Minimize Hamrds

Insufficient oxygen supply to Oxygen analyzer
the patient Pulse oximeter
Oxygen ratio monitor
Backup oxygen supply
Color-coded gas cylinders and hose
Touch-coded oxygen and nitrous oxide control knobs
DISS and pin-indexed safety system
Different diameters of hoses used
Vital sign monitoring
Insufficient carbon dioxide End-tidal carbon dioxide monitor
removal from the patient Pulse oximeter
Carbon dioxide absorber with color indicator
Excessive or wrong anesthetic Agent specific keyed filling spouts on the vaporizers
agent delivered to the patient with color coding
Interlock to prevent turning on more than one
anesthetic agent
Agent concentration monitors
BIS index monitor
Trauma to the lung caused by Pressure monitors
excessive pressure APL (adjustable pressure-limiting)valve
Pressure-limitingvalve
Foreign matters injuring the * Particle filters and traps
airway Dust-free carbon dioxide absorber
General * Regular service by qualified personnel (daily
functional check and regular quality assurance
inspection and maintenance)
-
 Chapter 34

DIALYSIS EQUIPMENT

OBJECTIVES

List the basic kidney functions and the functions of a hemodialyzer.

Describe the principles of diffusion, osmosis, and ultrafiltration.
Define ion/molecular clearance in a hemodialyzer.
Describe methods of vascular access and their advantages and disad-
vantages.
Analyze the construction of the artificial kidney (AK).
State the properties of the membrane in an AK and evaluate the mech-
anisms of molecular and fluid transport across the membrane.
Explain different methods of dialysate preparation and delivery in
hemodialysis.
Sketch line diagrams and identify the basic functional components in
the extracorporeal blood and dialysate delivery circuits.
Differentiate between peritoneal dialysis and hemodialysis.
Explain the needs and describe methods of water treatment in dialysis.
Compare different cleaning and disinfection methods for dialysis equip-
ment.

CHAPTER CONTENTS

1. Introduction
2. Basic Physical Principles
3. Kidney Functions Review
4. Mechanism of Dialysis
5. Hemodialysis System
 Biomedical Device Technolopy: Princ$les and Design

6. Dialyzer (or Artificial Kidney)

7. Patient Interface
8. Dialysate
9. Basic Components of a Typical Hemodialysis Machine
10. Peritoneal Dialysis
11. Other Medical Uses of Dialysis Treatment
12. Water Treatment
13. Cleaning and Disinfection

INTRODUCTION

A healthy kidney maintains the level of body fluid, electrolytes, and

acid/base balance, and removes some of the metabolic wastes. For patients
with impaired renal function, renal dialysis supplements or replaces some of
these functions to restore a reasonable state of health to the patient and min-
imize damage to other organs and physiological systems. The two most com-
monly used renal dialysis methods are hemodialysis and peritoneal dialysis.
The artificial kidney (AK) or dialyzer is a device that supplements or
replaces some of the many functions of the human kidney (e.g, water and
electrolyte balance, elimination of waste products, etc.). A very simplified
representation of the AK kinetics is shown in Figure 34-1. It consists of a
blood compartment and an electrolyte (or dialysate) compartment separated
by a semipermeable membrane.
Movement of substances, including water molecules across the mem-
brane, occurs in such an arrangement due to:

Blood to patlent

Figure 34-1. Artificial Kidney (Dialyzer).

 Dialysis Equipmant

Differences in concentrations of various substances in the two com-

partments
Particle sizes of these substances with respect to the membrane pores
Pressure difference between the two compartments
These phenomena are the result of diffusion and osmosis (or ultrafiltra-
tion), The procedure to use these physical principles in a controlled fashion
for medical purposes is known as hemodialysis.

BASIC PHYSICAL PRINCIPLES

Diffusion

Diffusion is the movement of molecules and ions in a solution as a result

of repeated intermolecular collisions. When two regions of a system have dif-
ferent concentrations, a concentration gradient is said to exist between the
regions. The rate of diffusion is proportional to the product of the concen-
tration gradient and the cross-sectional area separating the two regions.
In hemodialysis, the factors governing diffusion are complex because of
the presence of the membrane. The diffusion of substances (molecules) in a
solution through the membrane is substance-dependent as well as mem-
brane-dependent. The membrane may be characterized for a particular sub-
stance i by the permeability Pt of the membrane on the substance. The mass
diffusion rateJ of the substance i of the membrane is given by:

where C= the difference in concentration of the substance across the mem-

brane, and
I = the membrane thickness.
Note: Not all molecules can go through the membrane.

Osmosis

Osmosis takes place when a concentration gradient exists across a semi-

permeable membrane. A true semipermeable membrane is a membrane
through which only water molecules can pass. In dialysis, the membrane has
pores that allow certain sizes of molecules to go through. It is therefore not a
true semipermeable membrane.
The net movement of water across a membrane against a solute concen-
tration gradient is called osmosis. By means of osmosis, the system tends to
 Biomedical Device Technology: Princ@les and Design

achieve a situation of uniform concentration. This tendency for water to

move in response to a solute concentration gradient develops a force called
osmotic pressure. Osmotic pressure may also be defined as the hydrostatic
pressure that must be exerted on a solution to prevent the movement of
water through the semipermeable membrane. Osmotic pressure is a function
of the absolute temperature and the concentration gradient of substances in
the solution.

Ultrafiltration
If the opposing pressure to prevent movement of water through the
membrane is increased to a level above the osmotic pressure of the solution,
water is forced to flow from the solution against the osmotic pressure. This
event is called ultrafiltration. In hemodialysis, this can be used to remove
excess water from the patient. The mass of water transfer per unit time is pro-
portional to the pressure across the membrane.

KIDNEY FUNCTIONS REVIEW

Figure 34-2 shows the daily water transport of an average adult. About
2 liters of water is ingested per day; 200 ml of those is excreted from the
bowel and the rest is absorbed into the body. About 350 ml is lost to the
atmosphere during respiration in the form of water vapor in the expired air.
About 1,000 ml is excreted as urine through the urinary tract and 450 ml is
evaporated from the surface of the body. There are three water compart-
ments in the body: blood, intracellular (within the cells) fluid, and interstitial
(outside the cells) fluid.
An average person (70 kg) has about 40 liters of water (or about 40% by
weight) in the body. The percentage of water in a newborn is about 75%,
much higher than that of an adult. Of the 40 liters of water, 25 liters is with-
in the cells and 15 liters is in the interstitial fluid. There is about 5 liters of
blood in the body, of which 3 liters is plasma and 2 liters is red blood cells.
With a cardiac output of 5 liters per minute, about 1.2 l/min flows into the
renal arteries. The capillaries in the kidneys create about 2.2 m2 of mem-
brane contact surface area to process and filter the blood. On an average
day, 180 liters of fluid passes through the membrane inside the kidney but
almost all of it is reabsorbed, leaving only about 1 liter of fluid excreted as
urine.
 Dialysis Equipment

350 ml Lungs

1,000 ml
200 rnl
Figure 34-2. Body Fluid Transport.

Kidney Functions
The functions of a normal kidney include:
1. Removal of waste products from body fluid (blood and body water)
2. Regulation of blood volume
3. Regulation of extracellular fluid volume and composition
4. Maintenance of acid-base balance (pH)
5. Control of specific concentration of ions (e.g. Na, K balance)
6. Regulation of externally overtaken products such as glucose
7. Control of volume and composition of urine
8. Regulation of endocrine and metabolic functions
An artificial kidney (AK) replaces or supplements, to varying degrees, many
of these processes (except the last two functions).
One of the main functions of the kidney is the removal of waste products
from the blood. The parameter to measure the performance of the kidney in
terms of product removal is called plasma clearance. Similar parameters
called clearances are also used to evaluate a dialyzer's performance to
remove substances from the blood or bodily fluid.
Plasma clearance is substance-dependent. CLx (ml/min) of a substance x
is given by:
 Biomedical Device Technology: Principles and Design

where QU = rate of urine production (ml/min),

Cu = concentration of substance x in urine, and
Cpr= concentration of substance x in plasma.
Renal deficiency refers to underperformance of the renal system. Some
of the possible causes of renal deficiency are:
Kidney overload
Kidney disease or damaged kidneys
Inadequate renal blood flow
Some of the effects of renal deficiency include:
Uremia (increase in urea and other nonprotein nitrogen)
Water retention and edema
Acidosis in renal failure
Elevated K concentration in uremia
Uremic coma
Reduced availability of calcium to bones (osteomalacia)
Symptoms of renal failure include:
Loss of appetite, increased blood pressure, nausea, decreased urine
output, edema, itching, fatigue, and neurological disturbances. If left
untreated, will lead to convulsion, coma, and death.
One of the best means to assess renal failure is to measure concentra-
tions of the nitrogenous nonprotein substances (e.g., urea).
Treatment of renal problems includes:
Vigorous control of the intake of different materials into the body (diet
control)
Avoidance of excessive supply of water, cations, and nonessential
nitrogenous material
Hemodialysis treatment using an artificial kidney
Peritoneal dialysis: the use of the peritoneal membrane in the peri-
toneal cavity as the interface between blood and dialysis
Kidney transplantation

MECHANISM OF DIALYSIS

The dialyzer, also known as the artificial kidney (AK),is the main com-
ponent of the hemodialysis system in which blood solutes (metabolic wastes)
are removed from the blood and dialysate solutes (electrolytes) are added to
 Dialysis Equ$ment

the blood. During the process of dialysis, blood and dialysate are simultane-
ously circulated through the dialyzer, separated only by the semipermeable
membrane. Substances (including water) to be added to and removed from
the blood are exchanged across the membrane by the principles of diffusion
and ultrafiltration. The process usually continues for 4 to 5 hours per treat-
ment and consists of two to three treatments a week. The principles applica-
ble to dialysis are explained next.

Diffusion
Within a dialyzer, as the semipermeable membrane separating the blood
and dialysate is not an ideal membrane, ions (solutes) are selectively
exchanged across the membrane according to their molecular weight. Water
and waste products in the blood (e.g., urea, creatinine, uric acid), which have
relatively low molecular weights, can diffuse easily and rapidly through the
membrane into the dialysate; higher molecular weight substances, such as
glucose and proteins, cannot easily pass through the membrane and are not
significantly exchanged. Since the rate of diffusion depends on the concen-
tration gradients of the solutes, it is necessary to maintain a fresh supply of
blood and dialysate to facilitate these two-way transports.

Ultrafiltration

Ultrafiltration is the primary method to remove water from the blood

using the semipermeable membrane. In ultrafiltration, water is forced
through the membrane by applying a hydrostatic pressure across the mem-
brane. The physician may require intermittent adjustment of the ultrafiltra-
tion rate based on the observation and the long-term trend of the patient.

HEMODIALYSIS SYSTEM

Figure 34-3 shows a simplified functional block diagram of a hemodial-

ysis system. It consists of a vascular access serving as the interface between
the machine and the blood vessels of the patient. The blood circuit takes the
blood from the patient to the dialyzer and returns it to the patient. The blood
and dialysate are separated by the semipermeable membrane in the dialyz-
er where the substance and water exchange between the blood and dialysate
take place. The dialysate circuit takes the dialysate from the supply, feeds it
to the dialyzer and removes it from the dialyzer, and disposes it to the drain.
 Biomedical Device Technology: Princ$les and Design

Table 34-1 lists the functions described above plus other monitoring and
control tasks of each of the functional blocks.

Dialyzer
(AK)

I
Patient I
Blood I Dialysate
vascular
access circuit I supply
+ I
I

Figure 34-3. Hemodialysis System.

Table 34-1.
-
Functions of a Hemodialysis System
Vc~scularAccess -
Blood Circuit - --Dialyter
-
Dialysate
.-
Circuit Dialysate Supply
Remove blood Introduce and Provide Introduce and Prepare and
from patient remove blood blood/dialysate remove dialysate control dialysate
to dialyzer interface to dialyzer composition
Reintroduce Control/monitor Remove waste Control/monitor Remove air
blood to patient blood flow rate from blood dialysate flow rate from dialysate
Control/monitor Remove water Control/monitor Control/monitor
blood output from blood dialysate pressure dialysate
pressure temperature
Control/monitor Introduce solute Detect blood leak
blood input to blood into dialysate
pressure
Trap air bubbles Monitor dialysate
and produce pH
alarm
Prevent blood Monitor dialysate
clot conductivity
--- - --

DLALYZER (OR ARTIFICIAL KIDNEY)

The dialyzer (or AK) is the heart of the hemodialysis system where the
exchange of substances and removal of water between blood and dialysate
take place. The membrane within the AK allows such exchange to occur.
A membrane that permits substances to pass through is said to be per-
meable to those substances. A true or ideal semipermeable membrane is per-
meable to water but impermeable to all other substances. Most membranes
 Dialysis Equ$ment

pass only molecules of certain sizes. Hence they are called selective mem-
branes. Membrane permeability may be passive or active. Active perme-
ability transports molecules against the concentration gradient. Passive per-
meability depends on the concentration gradient as the driving force. The
ability of a particle to pass through a membrane passively is dependent on
the molecular size, ionic charge, and the degree of ionic hydration (e.g., OH-
is smaller and passes through the membrane easier than Ca++).
The basic properties of a dialyzer depend on:
Type of membrane used (porosity, size of pores, clearances, etc.)
Effective membrane surface area
Membrane's ability to withstand hydrostatic pressure
Transmembrane pressure
Blood flow rate
Dialysate flow rate

Performance Parameters
As discussed in the previous section, the parameter to measure the per-
formance of the kidney in terms of product removal is called plasma clear-
ance. A similar parameter known as clearance CL is also used to evaluate an
AK's performance to remove a substance x from the blood or bodily fluid.
In hemodialysis, CLx is redefined as:

C L = Q B X ( C a x - Cvx)
Cax

where QB = blood flow (ml/min),

Cax = arterial concentration of substance x, and
CU = venous concentration of substance x.
Another performance parameter of a hemodialysis system is the ultrafil-
tration coefficient (KUf). KUf is a concept used to evaluate an AK's perfor-
mance to remove water. KUf is defined as the volume (ml) of fluid that will
be transferred per unit time (hour) across the membrane per unit pressure
(mmHg) difference across the membrane of the dialyzer. Typical value of
KUf is 2 to 6 ml/hr/mmHg. However, it can be as high as 50 ml/hr/mmHg
for a high permeability dialyzer. Dialyzers with KUf greater than 8 ml/hr/
mmHg are generally referred to as high flux dialyzers.

Exampk
In a 4-hour dialysis treatment, it is necessary to remove 2 liters of water from
the patient. A dialyzer with KUf = 2.0 ml/hr/mmHg is used. It is estimated
 Biomedical Deuice Technology: Rinciples and Design

that during the treatment, 100 ml of fluid will be ingested by the patient. At
the end of the treatment, 300 ml of water will be used to rinse the dialyzer
free of blood (back into the patient). What is the pressure setting in the
dialysate compartment if the blood compartment has a position pressure of
50 mmHg?

Solution
The total water removal taking into account fluid ingestion and dialyzer rins-
ing is
(2,000 + 100 + 300) ml = 2,400 ml.
Let Xbe the pressure setting in the dialysate compartment. The transmem-
brane pressure Pis therefore equal to

Using the definition of KUE Volume of water removal = KUf X time X P,

and substituting the values into the equation:

to obtain a total transmembrane pressure of 300 mmHg.

Types of Dialyzers
Several types of AK with different physical constructions are available.
The more common types are coil, parallel plate, and hollow fiber. These AKs
are named according to the construction of the semipermeable membrane.
A coiled construction AK consists of a circular cross section tube made of
semipermeable membrane material wound into a coil. During dialysis, blood
flows inside the tube and the coil is immersed in a container filled with
dialysate. A parallel plate AK consists of multiple layers of semipermeable
membrane in parallel. Blood is circulated between alternate pairs of plates
and dialysate is circulated between the other plates. A hollow fiber AK con-
sists of a large number (10,000 to 15,000) of hollow fibers connected in par-
allel inside a container (Figure 34-4). Each fiber has an internal diameter of
about 0.2 mm and a length of about 150 mm. Blood flows inside the lumens
of the fibers with dialysate surrounding them. Although the lumen is small,
the diameter of an erythrocyte is 8 pm, a monocyte is 14 to 19 pm, and a
thombocyte is 2 to 4 pm. Hollow fiber AKs are the most popular type used
today. Common membrane materials are cellulose acetate, cuprophane,
nephrophane, and visking. The total surface area of the membrane ranges
from 0.6 to 2 m2 and supports a blood flow rate from 100 to 300 ml/min. A
 Dialysis Equipment

typical dialysate flow rate is between 400 and 600 m u m i n . Ultrafiltration

can be achieved by creating a positive pressure on the blood side or a nega-
tive pressure on the dialysate side.

f Used dialysate 4 Fresh dialysate

~ollow fibers in
dialysate container
Figure 34-4. Hollow Fiber Artificial Kidney.

Table 34-2 lists the construction and performance specifications of a hol-

low fiber artificial kidney for hemodialysis.

Table 34-2a.
Physical Specifications of AK.
Housing Construction Rigid transparent plastic
Tube Sheets Material Medical-grade silicon rubber
Dimensions 21 cm long x 7.0 cm diameter
Weight 650 g (filled)
Blood Volume 135 ml
Dialysate Volume 100 ml
Fiber Material Regenerated cellulose
Number of Fibers 11,000
Effective Length per Fiber 13.5 cm
Fiber Lumen 225 pm
Fiber Wall Thickness 30 pm
Effective Membrane Area 1.0 m"
-- --

Dialyzers are evaluated by comparing:

Clearances for different substances (e.g., urea, creatinine, phosphate,
vitamin Biz, uric acid, blood serum phosphate, glucose, NaCl)
Ultrafiltration rate (KUf)
Priming volumes
 Biomedical Device Technology: Pfinc$les and Design

Table 34-2b.
Performance Specifications of AK.
Rlood Compartment At blood flow rate of 200 ml/min 15 to 55 mmHg
Flow Resistance
Dialysate Compartment At dialysate flow rate of 500 ml/min with
Flow Resistance negative pressure of 400 mmHg
Average Ultrafdtration At 500 ml/min dialysate flow and
Rate 300 mmHg negative pressure
Typical Urea At 200 ml/min blood flow and
Clearances Creatinine 500 ml/min dialysate flow
Phosphate

Cost (taking into account single or multiple use)

Clotting properties

PATIENT INTERFACE

As hemodialysis is often done three times a week, a semipermanent

interface between the dialysis unit and the patient's circulatory system to
allow repeated dialysis is required. This interface is called the vascular
access. Several common methods to access the bloodstream are described in
the following sections.

A-V Shunt
An A-V shunt is a pair of cannulae of polytetrafluoroethylene (FTFE)
inserted through the skin into an artery and a vein near the inner surface of
the forearm or the lower leg. Between dialysis treatments the two cannulae,
which are permanently implanted, are joined by a short length of SilasticTM
tubing to allow blood circulation. During dialysis, the SilasticTMtubing is
removed and replaced by two plastic tubings that direct the blood to and
from the dialyzer. These A-V connections can provide natural blood pressure
differential to circulate blood through the dialyzer. However, due to the low
differential pressure, it requires a low flow resistance dialyzer; otherwise a
pumping mechanism is necessary to provide enough blood flow. A-V shunts
are used in acute as well as chronic therapies.
 Dialysis Equ$ment

A-V Fistula
In this method, an internal shunt is developed by joining an artery and a
vein within the limb by a short length of fibrin tubing. Blood is obtained by
venous puncture using either one or two large-bore needles. A blood pump
is necessary to create enough blood flow.
In the double-needle technique, blood is continuously and simultane-
ously withdrawn from one needle and reinfused through the other. The sin-
gle-needle technique requires a Y-connection and a controller to alternately
withdraw and infuse blood to and from the patient. A special pump or a pair
of synchronized pumps is required for this technique. To provide continuous
blood transfer, a special double-lumen needle/catheter can be used.
A-V fistula is the most commonly used method for vascular access as it
requires no permanent open site, which minimizes infection problems
encountered with a normal A-V shunt. It has a 3-year 70% average site sur-
vival rate.

A-V Graft
Similar to an A-V fistula, this vascular access method surgically puts in a
graft (a section of autogenous saphenous vein or PTFE TeflonTM) to connect
an artery to a vein (e.g., graft between the radial artery and basilic vein). It
has a 3-year 30% average site survival rate.

Percutaneous Venous Cannula

This method is only for acute cases. To create the vascular access, a can-
nula can be inserted into the subclavian, femoral, or internal jugular vein. A
percutaneous venous cannula can be single or double lumen.

DIALYSATE

Dialysate is a solution to effect diffusion and ultrafiltration. It is also used

to carry away waste products and neutralize excess body acid by transferring
either acetate or bicarbonate from the dialysis solution. There are two basic
types of dialysate: acetate base and bicarbonate base. Dialysate is prepared
in batch before dialysis or on demand during dialysis. Batch processing is
done only in large dialysis centers where the dialysate is prepared and stored
in large holding tanks. In a smaller center or when prescription dialysis is
required, the dialysate is prepared by blending dialysate concentrate in pro-
 Biomedical Device Technology: Princ@lesand Design

portion with treated water during the dialysis process. The concentration
must be continuously monitored using thermally compensated conductivity
meter. The mechanism for dialysate blending, monitoring, and control is
often an integral part of the dialysis machine. Examples of dialysate compo-
sitions are listed in Table 34-3. The unit of concentration is rnEq/l.

Table 34-3.
- -
Dialysate
---
Compositions. - -

Nu
-
K
-
Ca M g - -
Cl Acetate Dextrose
Standard 134 2.6 2.5 1.5 104.0 36.6 2.5
K Free 134 0 2.5 1.5 101.0 37.0 2.5
Low Ca 2.6 1.0 1.5 102.5
Low Na 130 2.0 3.0 1.0 101.0
- - --- -

In any case, the following processes are necessary before the dialysate is
introduced into the AK or during dialysis:
Treat water before it is used to prepare the dialysate (water treatment
is discussed in the latter part of this chapter)
Warm the dialysate to body temperature (37OC) before entry into the
dialyzer
Deaerate the dialysate to prevent gas evolution at body temperature
and subatmospheric pressure (for AK using negative pressure ultrafil-
tration) and to prevent supersaturation of blood with nitrogen at body
temperature
Monitor dialysate pressure at entry to and exit from the dialyzer to
ensure that the blood pressure is always greater than that of the
dialysate
Detect blood leaks across the membrane using photoelectric or ultra-
sound detector
In modern day dialysis, the dialysate is continuously fed into and
removed from the A K The used dialysate is disposed of during dialysis. This
is referred to as a single-pass dialysate system. In earlier day dialysis, in order
to conserve chemicals (and cost), dialysate was reused on the same patient.
Many methods to reuse dialysate have been employed. For example, sorbent
materials are used to remove some chemicals from the dialysate so that it can
be regenerated and reintroduced into the AK (sorbent regenerative system).
In another method, the used dialysate may be mixed with a certain propor-
tion of fresh dialysate and reintroduced into the AK (single-passrecirculation
system). In the extreme case, the used dialysate is recirculated back into the
 Dialysis Equt$ment

AK until the performance of dialysis has decreased to such a level that the
old dialysate must be replaced with a fresh batch (total recirculating system).

BASIC COMPONENTS OF A
TYPICAL HEMODIALYSIS MACHINE

Figure 34-5 shows the basic components and fluid flow diagram of a typ-
ical hemodialysis machine. Blood enters the machine from the vascular
access via the arterial blood line, through the artificial kidney (dialyzer),and
returns to the patient via the venous blood line. Dialysate is prepared and fed
into the AK, where water and substance exchange take place. After passing
through the AK, the dialysate is dumped into the drain. The blood circuit is
separated from the dialysate circuit. Blood is separated from the dialysate as
long as the membrane in the AK remains intact. The basic components in
the extracorporeal blood and dialysate delivery circuits are described in the
following sections.

(Extracorporeal Blood Circuit) ,Blood pump

.....,......Heat
.........,........,..... Blood leak
exchanger Flow sensor detector

To drain

Figure 34-5. Hernodialysis Machine Basic Components and Flow Diagram.

 Biomedical Device Technology: Princ$les and Design

Extracorporeal Blood Circuit

Heparin is administered intermittently or continuously (e.g., at a rate of
1.5 mVhr) to the incoming blood from the venous access. Blood clots
decrease the efficiency of the AK and are hazardous if they are reintroduced
into the patient. Regional heparinization is used in some clinical conditions
( e g , to avoid worsening of existing bleeding sites). During regional hepari-
nization, careful neutralization of heparin is achieved by injecting coagulants
(such as protamine) into blood returning to the patient.
A roller pump in the dialysis machine controls the blood flow rate and
creates a positive pressure in the blood compartment of the AK. The blood
flow rate is controlled by varying the speed of the pump. The pressure in the
blood line is regulated by the roller pump and the venous clamp located at
the end of the blood circuit in the machine. The positive pressure in the
blood circuit also prevents ingress of air and dialysate fluid into the blood cir-
cuit.
An arterial drip chamber provides a visual indication of the blood flow
in the circuit. The pressure and flow rate of blood before entering the AK are
monitored. As the blood exits the AK, it passes through another drip cham-
ber. As the pressure becomes less, foaming occurs as air exits the blood. To
prevent excessive air from getting into the bloodstream of the patient via the
returning blood, an aidfoam detector is located at the drip chamber. A pres-
sure sensor monitors the blood pressure at the exit of the AK. Together with
the upstream pressure sensor, the sensors monitor the blood average com-
partment pressure in the AK for ultrafiltration control as well as measuring
the resistance of the AK. If the hollow fibers are blocked by blood clots, a
large pressure drop will develop across the AK in the blood circuit.

Dialysate Delivery Circuit

Water from the purified water source enters the dialysate circuit of the
dialysis machine. It is first heated when passing through the heat exchanger
and the heater compartment. A metering pump accurately introduces con-
centrated dialysate into the water path to produce the right concentration of
dialysate to the patient. Gas in the dialysate is removed by passing the
dialysate through the deaeration chamber, where a reduced pressure is cre-
ated by the deaeration pump. Excessive gas in the dialysate can diffuse
across the membrane into the patient's blood.
Temperature, pressure, flow rate, pH, and conductivity of the dialysate
are measured in the metering chamber. To prevent cooling or heating the
patient's blood, the temperature sensor ensures that the dialysate is at body
temperature before entering the AK. Correct flow rate and pressure are
 Dialysis Equ$ment

required to ensure removal of substances and water from the patient. The
dialysate pH and conductivity are indications of the dialysate composition
and concentration.
After passing through the metering chamber, the dialysate is introduced
via a check valve (one-way valve) into the AK. A dialysate pump maintains
the flow rate and produces a negative pressure in the dialysate chamber of
the AK. The positive pressure in the blood circuit and the negative pressure
in the dialysate circuit create the transmembrane pressure that controls the
ultrafiltration rate, whereas the blood flow and dialysate flow rates control
the substance removal rate (clearances) of dialysis. Before going through the
heat exchanger and being dumped into the drain, the dialysate passes
through a blood leak detector. If blood is detected in the dialysate, which
indicates rupture of the membrane in the AK, the machine will sound an
alarm and have to be shut down. The dialysate bypass line can be used to
facilitate replacement of the AK.
Table 34-4 shows the typical range of control and monitoring parame-
ters of a hemodialysis machine.

Table 34-4.
Range of Control and Monitoring Parameters.
- --

Parameters Typical Range

Dialysate flow rate 200-1,000 mVmin
Dialysate temperature 35-3Y°C
Conductivity 7-17 mS/cm
Blood flow rate
Heparin flow rate
Venous/arterial pressure display -300 to +600 mmHg
Transmembrane pressure -100 to +500 mmHg
Blood leak detector sensitivity 0.35-0.45 mVmin
Air detector sensitivity Air bubble size > 5-25 pl in blood line
Ultrafiltration rate 0-4 1/hr
-

PERITONEAL DIALYSIS

Hemodialysis requires removing blood from the patient and processing

the blood in the AK external to the patient's body. Another form of dialysis
can be performed using a natural membrane inside the human body to
achieve substance and water exchange. As the peritoneal cavity is lined with
blood vessels and capillaries, peritoneal dialysis uses the peritoneal mem-
brane as the blood-dialysate interface instead of a membrane in an external
 Biomedical Device Technology: Principles and Design

dialyzer. A peritoneal dialysis setup is much simpler than a hemodialysis sys-

tem. It uses gravity feed and drain instead of blood and dialysate pumps. The
dialysate stays in the peritoneal cavity instead of being circulated through the
external AK. Figure 34-6 shows a setup for continuous cycler-assisted peri-
toneal dialysis (CCPD).

Dialysate
supply

Heater and

control

Patient's
peritoneal
Drain bag cavity
with scale

Disposal
bag

Figure 34-6. Continuous Cycler-Assisted Peritoneal Dialysis Setup.

At the start of a dialysis cycle, valve number 4 opens (all others remain
shut) to allow a selected volume of dialysate to flow from the supply reser-
voir to the volume control and heater compartment. The dialysate stays in
the compartment until it is warmed to body temperature. Valve number 2 is
then opened so that the dialysate flows by gravity to the patient's peritoneal
cavity through an indwelling catheter. The dialysate stays inside the peri-
toneal cavity for a period of time ( e g , 45 minutes) to allow substances and
water exchange between the blood in the capillaries and the dialysate. After
the preset time, valve number 3 opens to drain the used dialysate (together
with the additional water from osmosis) from the peritoneal cavity to the first
drain bag. The scale measures the weight of the dialysate to monitor the fluid
removed from the patient. After the measurement, valve number 1 is opened
to allow the dialysate to flow into the disposal bag to complete the cycle.
Although peritoneal dialysis takes more time due to slower transport, the
 Dialysis Equipment

rate and process have more resemblance to those of natural kidneys and
therefore reduce the possibility of shock to the patient. Peritoneal dialysis is
often performed at home due to its relatively simple operation and less
sophisticated equipment setup. However, because of the high risk of devel-
oping peritonitis (due to careless handling of indwelling catheters by patients
or home caregivers), patients are often forced to switch to hemodialysis due
to repeated peritonitis. Continuous ambulatory peritoneal dialysis (CAPD)
and continuous cycler-assisted peritoneal dialysis (CCPD) are the two com-
monly performed types of peritoneal dialysis.

Continuous Ambulatory Peritoneal Dialysis (CAPD)

In CAPD, the dialysate is constantly present in the abdomen but is
changed three to five times daily with a per-fill-volume from 1.5 to 3.0 liters
(typically2 liters). Dextrose (1.5, 2.5, or 4.25%) in the dialysate is used to cre-
ate osmosis for water removal. Drainage and replenishment of dialysate are
performed using gravity.

Continuous Cycler-Assisted Peritoneal Dialysis (CCPD)

CCPD is performed at bedtime using an automated cycler to change
dialysate four to five times during the night. The dialysate used is similar to
that used in CAPD.

OTHER MEDICAL USES OF DIALYSIS TREATMENT

Other than treating patients with renal problems, dialysis may be used to
eliminate toxic materials in the blood, to perfuse isolated organs, to reduce
abnormally high ammonia concentration found in the blood following liver
malfunction, or to supplement renal function during and after major surgery.

WATER TREATMENT

Normal tap water contains traces of metal ions (e.g., copper, lead) and
chemicals (e.g., chlorine or chloramines). During a 4-hour dialysis treatment
using a dialysate flow of 500 ml/min., 120 liters of dialysate interface with the
patient's blood. Under repeated dialysis, if untreated water is used to prepare
the dialysate, these ions and chemicals, which normally are not harmful
 Biomedical Device Technology: Principles and Design

under usual consumption, can quickly accumulate in a patient's body.

Therefore, it is necessary to remove ions (e.g., Cut+) and other molecules
(suspended particles) in the water used for dialysate preparation.
Raw water usually goes through a pretreatment process that consists of:
1. Cartridge filter to remove particles greater than 0.05 pm (but cannot
remove dissolved toxin or endotoxin)
2. Water softener to remove ions such as Ca++or Mg++
3. Activated carbon to remove chlorine or chloramines
A second-stage water treatment process is carried out to remove smaller
particles and remaining chemicals in the water. Reverse osmosis (RO) is the
most commonly used method. Distillation can also be used. RO (using the
same principle as ultrafiltration) is achieved by applying pressure, forcing
water to pass through a true semipermeable membrane while leaving impu-
rities behind. RO removes over 90% of impurities and is acceptable in most
cases.
To further purify the water, ion exchange resins can be used to remove
all charged ions from the water. Activated charcoal is used to remove non-
toxic contaminants. Heat or ultraviolet radiation can be used for sterilization.
However, in most cases, there is no need for water sterilization since the
membrane in RO or distillation can remove the majority of bacteria and
endotoxins. For dialysis applications, water with a bacteria count below 200
colonies per ml is acceptable.

CLEANING AND DISINFECTION

As dialysis is an invasive procedure, care must be taken to reduce the

chance of infection. Dialysis machines are required to be flushed and disin-
fected between patients. Heat disinfection (e.g., heat fluid to 85OC for 15 min)
can eliminate waterborne bacteria such as Pseudomonas cepacia (a gram-
negative bacterium). Although heat disinfection is convenient, it is ineffective
to kill spore-forming bacteria such as Bacillus varieties. Chemical disinfec-
tion (e.g., using formaldehyde) can be used instead of heat treatment.
Disinfection treatment using heat or chemicals is usually done daily on every
machine and after each patient's treatment.
In addition to daily disinfection, sodium hypochorite (bleach) is used to
disinfect the machine weekly. It is important to rinse the machine thorough-
ly after chemical or bleach disinfection to remove all chemical residuals
before the machine is used on patients.
A dialysis center must set up standard operation procedures to take reg-
ular bacteria cultures in order to monitor the effectiveness of disinfection and
 Dialysis Equ@ment

detect possible contamination.

The blood and dialysate lines are single use disposable items. Although
all dialyzers are labeled single-use, many centers reuse a dialyzer on the
same patient. Studies have shown that following proper cleaning and disin-
fection procedures, reusing a dialyzer on the same patient is easier on the
patient (i.e., it has less chance to cause adverse reactions). Some centers have
shown that after taking into account the time and materials for processing,
reusing dialyzers can achieve cost saving.
 Chapter 35

MEDICAL LASERS

OBJECTIVES

Describe the characteristics of lasers and their applications.

Explain the effect of lasers on tissue.
Describe the physics of laser action.
List different surgical lasers and their applications.
Explain the two common methods of laser delivery.
Identify the benefits and limitations of laser surgery over conventional
surgical methods.
List the hazards associated with the use of lasers and the methods to
mitigate the risks.
State the maintenance requirements and handling precautions of laser
systems.

CHAPTER CONTENTS

1. Introduction
2. Characteristics of Lasers
3. Laser Action
4. Applications of Lasers
5. Tissue Effects and Surgical Applications
6. Characteristics of Medical Lasers
7. Functional Components of a Surgical Laser
8. Laser Delivery System
9. Advantages and Disadvantages of Laser Surgery
10. Safety Hazards and Risk Mitigation
11. Maintenance Requirements and Handling Precautions
 Medical Lasers

INTRODUCTION

Laser is the acronym for Light Amplification by Stimulated Emission of

Radiation. Albert Einstein in 1917 described the absorption, spontaneous
emission, and stimulated emission of light, which eventually lead to the
development of lasers in 1940 in the former Soviet Union.
An electron, at its resting or ground state, when struck by a photon, can
absorb the energy of the photon, become excited, and move to a higher
energy level. On spontaneous return to its ground energy level, the electron
emits a photon of energy equal to the difference between the two energy lev-
els. This emitted photon can interact with an atom with an excited electron
to produce another photon with the same frequency and phase traveling in
the same direction. When there are many excited atoms in the medium
(known as having a high degree of population inversion), this mechanism
will set up a chain reaction of stimulated emission. Stimulated emission
under the right conditions creates light amplification.
The first laser used in surgery was a ruby laser for treatment of retinal
hemorrhages in the United States in the 1960s. It was not until 1972, when
Gezo Jako adapted a COz laser to an operating microscope, that laser
became a viable surgical instrument in operating rooms. This chapter dis-
cusses some of the applications of lasers in medicine.

CHARACTERISTICS O F LASERS

Although both are electromagnetic waves, a laser is different from a nor-

mal light source. Lasers have the following characteristics:
Monochromatic-Lasers have one or a few discrete wavelengths due
to the fixed energy band gaps of the atoms, whereas normal light consists of
a relatively wide spectrum of wavelengths. However, in practice, a laser has
a finite (but narrow) width of wavelength (Figure 35-la).
Coherent-Due to stimulated emission, the waves or photons coming
from the laser are all in phase, whereas those from normal lights have dif-
ferent phase angles (Figure 35-lb).
Collimated-Due to the repeated reflection between the parallel mir-
rors (generation of laser is discussed in the latter part of this chapter), all the
waves of the laser beam are parallel along the longitudinal axis of the mir-
rors. Compared to normal light, the trajectory of the laser beam coming from
the lasing medium has minimal divergence or convergence (Figure 35-lc).
 Biomedical Device Technology: P/inc$les and Design

Laser In-phase
Normal
light I /wave'eis Laser beam

I
Wavelength
(a) Monochromatic (b) Coherent (c) Collimated
Figure 35-1. Laser Characteristics.

LASER ACTION

Although there are many types of lasers, they all have three basic com-
ponents:
I. A lasing medium, which can be gas, liquid, or solid
2. An external excitation source that pumps energy into the lasing
medium
3. A resonator or optical cavity with two parallel mirrors housing the
lasing medium. One mirror is totally reflective and the other is par-
tially reflective
Using a ruby laser as an example, it consists of a flash lamp (excitation
source), a ruby crystal (lasing medium), and two mirrors as shown in Figure
35-2. The flash lamp ignites and pumps energy into the ruby atoms. Light
energy is absorbed by the atoms in the ruby crystal to excite electrons to
higher energy levels. Some of the excited electrons return to their ground
state and emit photons. The photons traveling in the direction perpendicular
to the mirrors are bounced back and forth between the two mirrors. As they
travel inside the crystals, they stimulate more photon emissions from the
excited atoms. The beam intensity therefore increases as it undergoes multi-
ple reflections along the longitudinal axis between the mirrors. A portion of
the beam is allowed to leave the laser through the partially reflective mirror.

APPLICATIONS OF LASERS

Lasers are used in a wide range of products and technologies.

Applications of lasers can be separated into three categories.
 Medical Lasers

Flashbulb
Llght energy
pumped ~ n t olasl L~ghtphoton
med~um

AcOve laser

Total reflective Ruby ctystal Partial reflective

mirror (lasing medium) mirror
Figure 35-2. Laser Action.

Thermal Effect
When a laser is absorbed by a target, it converts to heat energy. A lens
system or a light pipe can be used to focus and redirect a laser. This proper-
ty is used in the industry in cutting materials such as metal or burning a com-
pact disc. In the military, its heating effect is used in laser guns to destroy mil-
itary targets. In medicine, lasers are used in surgery and physiotherapy.

Straight Collimated Beam

The collimated property of a laser beam produces a parallel beam of
light that has little convergence and divergence. That is, the beam diameter
of an ideal laser will stay constant irrespective of the distance. Due to this
property, a laser beam can travel a long distance without losing its intensity
(except from absorption in the optical path). Lasers are widely used in indus-
try such as in land survey and in precision alignment. In the military, laser
beams are used in guidance systems such as laser-guided missiles. In medi-
cine, this property is used in position alignment such as beam alignment of
linear accelerators in cancer treatment.

Photostimulation Effect
As a laser produces a monochromatic beam of high-intensity light, it can
be used as a stimulant. In medical applications, a laser beam can be used to
stimulate blood circulation and in cell healing in physiotherapy. In addition,
it can be used in conjunction with a photodynamic drug to selectively acti-
vate the drug by a laser beam.
 Biomedical Device Technology:Princ$les and Design

TISSUE EFFECTS A N D SURGICAL APPLICATIONS

The surgical effect of a laser is primarily due to its thermal effect on tis-
sues. Laser tissue effect depends on:
1. Type of tissue
2. Type of laser
3. Power density at the lasing site
4. Exposure time (continuous/pulsed)
The general tissue effect inflicted by a surgical laser is shown in Figure
35-3. In essence, when the laser beam hits the tissue, the laser energy is
absorbed by the tissue to create three zones of injury. Due to the intense heat,
the cell membranes rupture and vaporize at the center of the laser beam
(zone 1).Next to the vaporized zone is a zone of cell necrosis where the tis-
sues undergo irreversible heat damage (zone 2). Beyond the necrosis zone is
a layer of cells that were injured due to the elevated temperature (zone 3).
Tissues in zone 3 are able to repair and recover.

Laser beam

Healthy tissue
Necrosis

Figure 35-3. Zones of Injury from a Surgical Laser.

The rise in temperature and temperature distribution in the tissue during

laser irradiation depends on the energy absorbed and the thermal parameter
of the tissue. For short irradiation (less than a few seconds), tissue heated to
less than 60°C undergoes little or no permanent damage. Denaturation of
protein will result when the tissue temperature is above 60°C. Coagulation
of blood occurs when the blood temperature is above 82OC. Tissue charring
and vaporization starts to occur above 90°C. The heat absorption effect of
 Medical Lasers

tissue when hit by a laser depends on the type of tissue and the type of laser.
For examples, an argon laser is highly absorbed by hemoglobin but not by
water. Using this property, an argon laser can be used as a photocoagulator
to stop bleeding at the back of the eye. In the procedure, the laser beam pass-
es through the cornea with very little or no absorption and delivers its ener-
gy to the blood vessels on the retina. On the other hand, a CO:! laser is high-
ly absorbed by water, which makes it a general surgical laser as all soft tis-
sues contain a high percentage of water.
The power density or intensity of a laser beam striking an object is equal
to the power divided by the beam area on the object. A laser, like light, can
be reflected by a mirror or focused by a lens. A laser beam can be focused
to a tiny spot to produce a very high intensity beam. Figure 35-4 shows the
tissue effects of different focal spot size of the same laser. In general, the high-
er the beam intensity, the deeper the vaporization zone.
For a continuous laser, the longer the exposure time, the more energy the
tissue will absorb; in terms of a surgical laser, the deeper and wider the zone
of injury will be. A laser can be pulsed to increase its peak power, creating
intense heat in short durations. A pulsed laser creates a deep zone of vapor-
ization but allows periods of cooling between pulses. Such cooling periods
slow heat conduction to adjacent tissues. A pulsed laser will provide a deep-
er vaporization with less surrounding tissue damage than continuous laser at
the same power output. By manipulating these parameters, different surgical
effects can be created.
When an intense beam of laser slides across the surface of a soft tissue, it
produces a zone of vaporization along the path of the laser. This action pro-

Focus~nglens Converging
laser beam
I I 1
I I I
I I
I I
I I
I I -I
I I I
I I
I I I
I I I
I I I

Parallel laser
beam

Figure 35-4. Tissue Effect on Power Density (or Focal Spot Size).
 Biomedical Device Technology: Pfinc$les and Design

duces a sharp, clean cut. Abrasion effect (removal of a thin layer of surface
tissue) is created by moving a defocused beam of laser with sufficient inten-
sity over the tissue. Laser energy absorbed by the tissue may create destruc-
tion and charring effects on the tissue. Absorption of laser energy by blood
produces coagulation effect. Retina reattachment, vision correction, vascular
surgery, and microsurgery are some of the many examples of laser surgical
applications.

CHARACTERISTICS OF MEDICAL LASERS

There are many types of lasers. Lasers are often classified by their lasing
media. A solid laser has lasing material in solid crystal form such as a ruby
crystal. In gas lasers, the lasing media are gas mixtures. Examples of gas
lasers are helium neon (HeNe) and carbon dioxide (COz) lasers. The name
excirner h e r is derived from "excited and dimmers." A reactive gas mixture
is electrically stimulated to form a pseudomolecule (dimer) and when excit-
ed produces a cool laser. A dye laser uses a fluorescent liquid dye as the las-
ing medium. When exposed to an intense laser such as an argon beam, it
absorbs the laser energy and fluoresces over a broad spectrum. A tunable
prism can be used to adjust the wavelength. A semiconductor diode can be
manufactured to emit laser. However, the output power of a semiconductor
laser is usually too low for surgical applications. Table 35-1 lists the charac-
teristics and applications of common surgical lasers.
Transverse electromagnetic modes (TEM) of the laser beam are due to
the oscillatory behavior of the electric and magnetic fields at the boundary
of the laser resonator. The shape of the transverse mode is shown by the
shape of the output beam. Figure 35-5 shows the shapes of selected trans-
verse laser modes. These modes can be visualized by the burn mark on a
wooden tongue blade by irradiating the laser beam vertically on the tongue
blade. The fundamental mode of TEMoo with a Gaussian beam intensity pro-
file is the best mode structure to maximize the energy density propagation.

FUNCTIONAL COMPONENTS OF A SURGICAL LASER

Figure 35-6 shows the functional components of a surgical laser. A low

power helium neon (HeNe) laser producing red light is often used as the
"aiming beam." An alignment optical system is used to align the HeNe beam
with the main laser beam. Most surgical lasers require high power output and
 Medical Lasers

Table 35-1.
Laser Characteristics and Applications. -- --

-
her
-
Wavelength Color Lasing Medium
- -
Applications -

C02 10.6 pm Far- Mixture of carbon Absorbed by water. For

infrared dioxide, nitrogen, vaporization and cutting tissue.
and helium gas
Holmium: 2.1 pm Mid- Crystal of holmium, Absorbed by tissue containing
YAG infrared thulium, and water. Precise cutting and less
chromium generalized heating of tissue
ND: YAG 1.064 pm Near- Crystal of Poorly absorbed by hemoglobin
infrared neodymium, and water, but absorbed by
yttrium, aluminum, protein. For denaturation of
and garnet protein and shrinkage of tissue
and coagulation
Ruby Red Ruby crystal Not absorbed by blood vessels
and transparent tissues. High
energy pulses selectively vaporize
tissue, use in dermatology and
plastic surgery such as port-wine
stain removal
HeNe Red Helium neon gas Use as aiming beam of medical
lasers
Green Crystal of potassium Highly absorbed by red or dark
titanyl phosphate tissue. Use for coagulation and
precision work
Argon 514 nm Green Argon gas Passes through water and clear
fluid but highly absorbed by
488 nm Blue red-brown pigments. Use in
coagulation, ophthalmology,
dermatology, and plastic surgery
Excimer 193 nm Ultraviolet Argon fluoride gas Precision cutting and coagulation
with little thermal damage to
308 nm Xenon chloride gas surrounding tissue. Use in
ophthalmology, angioplasty,
351 nm Xenon fluoride gas orthopedics, and neurosurgery
Dye- 400-900 Entire Fluorescent liquid Use in photodynamic therapy,
tunable nm visible dyes dermatology, and plastic surgery
spectrum
- --

therefore a cooling system must be in place to take away the heat generated
from the laser production. The laser will then be coupled to the surgical site
via a system of delivery device (or transport medium).
 Biomedical Device Technology:Princ$les and Design

1tMii
Figure 35-5. Transverse Electromagnetic Modes.

- ' Excitation1
energy pump
Cooling
u +
2$ Laser output
"s 2
+ Laser generator 0 6
u n, 3 Delivery 2
o_ 8: device to tissue
Y
HeNe laser (aiming beam) 2

Figure 35-6. Functional Components of a Surgical Laser.

LASER DELIVERY SYSTEM

A laser procedure may be contact or noncontact. In noncontact proce-

dures, the surgical delivery of the laser to the surgical site can be via a sys-
tem of mirrors or through an optical fiber. Some lasers (such as Nd:YAG,
Argon) can be delivered using optical fibers while others (far-infrared lasers
such as Con) can be delivered only through a system of mirrors. Figure
35-7a shows an articulation arm system for CO2 laser delivery. The laser
travels inside a hollow rigid tube and is reflected to another tube segment by
a first surface mirror located at the junction of the two tube segments. After
 Medical Lasers

Laser output Rigid arm

Articulated joint
\ /

Laser beam
(a) Laser Articulation Arm and Mirror Optics

Laser beam

(b) Laser Optical Fiber Showing Total Internal Reflect~on

Figure 35-7. Laser Delivery Systems.

several reflections, the laser will reach the handpiece and can be directed at
the surgical site.
Instead of using mirrors, the laser may travel inside a flexible optical
fiber by total internal reflection (Figure 35-713). The laser travels inside the
optical fiber until it has reached the other end of the fiber. Due to its small
diameter and flexibility, optical fiber can easily move around the surgical
sites. Silica fibers are good for ultraviolet and visible lasers and glass fibers
are good for visible lasers. Special fibers are required for near- and mid-
infrared lasers.
Contact laser probes offer a completely different method of delivering
laser energy to tissue. Instead of laser directly transferring its energy to the
tissue, the laser first heats the contact probe, and the heat of the probe in con-
tact with the tissue is used to create the surgical effect. Similar to noncontact
lasers, contact laser probes will cut, coagulate, vaporize, and ablate. These
probes can be attached to a variety of handles for use in open surgical pro-
cedures or can be affixed to a standard optical fiber and passed through any
rigid or flexible endoscope. Contact laser probes can be manufactured to any
shape and are made of synthetic sapphire crystals with great mechanical
 Biomedical Device Technology: Pfinc$les and Design

strength, low thermal conductivity, and high melting temperature. The tip of
a probe can be heated to 2,000°C.

ADVANTAGES AND DISADVANTAGES OF LASER SURGERY

Lasers are selectively absorbed by different tissues to produce different

surgical effects. Lasers offer many advantages over other surgical techniques.
The laser beam can be precisely focused for localized destruction of
tissue. The depth of penetration can also be regulated by the power
density, duty cycle, duration, and focus.
Using fiber optics, a laser can access areas that are inaccessible to other
surgical instruments.
The laser beam simultaneously cuts and coagulates blood vessels and
seals lymphatic vessels.
Less tissue trauma due to no pressure and no traction applied on tis-
sue.
Quick healing with minimal postoperative edema, little pain, and lit-
tle scarring.
Noncontact procedure reduces risk of contamination and infection.
Some of the disadvantages of laser surgery are:
Safety risks to patient and staff in terms of potential eye injuries, burns,
and fire hazards.
Higher cost per procedure due to expensive equipment and setup.
Need for special facility support, supplies, and staff training.

SAFETY HAZARDS AND RISK MITIGATION

Surgical lasers present hazards to patients and to clinicians, Their use

must comply with regulations, standards, manufacturers' recommendations,
and professional practices. Lasers must be operated only by trained users, in
designated areas, and adhering to institutional policies and procedures. To
ensure laser safety in a surgical procedure, all users must be familiar with the
specific laser to be used, its accessories, modes of operations, tissue effects,
and potential risks.
 Medical Lasers

Eye Protection
In addition to hazards common to all electromedical devices, a high-
energy laser beam (with its collimated property) can cause damage at a dis-
tance far from its source. Inadvertent firing of a laser may cause burns on
patient or staff, ignite a fire, or even cause an explosion in an oxygen-
enriched environment. Many lasers are in the infrared or ultraviolet range in
the electromagnetic spectrum, which are invisible to the human eye. An
operator may not be aware of the laser path until damage is done. When a
laser beam is directed to the eye, the collimated beam of a laser will be
focused by the lens to a small dot of high-energy density on the back of the
eye. This high-intensity beam will create irreversible damage to the eye.
Even a low-energy laser beam, which normally will not create tissue burns,
will have enough power density to inflict ocular injuries after being focused
by the lens of the eyes. An acute exposure to laser can cause a scotoma (per-
manent damage of a small area of the retina), resulting in a blind spot in the
field of vision. Long-term exposure to low-energy laser may lead to slow
degenerative changes due to thermal or photochemical injuries. Examples of
such injures are slow cataract formation in damaged lens and chronic reduc-
tion of color-contrast sensitivity from a damaged retina.

Laser Classifications
Based on these potential hazards, lasers are classified according to their
risks, especially in ocular exposure. According to these classifications, safety
measures and special precautions are required during laser procedures. The
following classifications are taken from the Standards "CAN/CSA-Z386-92-
Laser Safety in Health Care Facilities." Similar definitions are found in
"ANSI 2136.3-1988-Standards for the Safe Use of Lasers in Health Care
Facilities." The Standards also stipulate responsibilities of health care facili-
ties; composition and responsibilities of laser safety committee and the laser
safety officer; safety control measures; risk management and quality assur-
ance guidelines; training, education, and credentialing of laser users.
Class I. Laser equipment emitting radiation that is not considered haz-
ardous. These lasers do not require hazard-warning labeling.
Class IIa. Laser equipment emitting radiation that is not considered haz-
ardous when viewed for up to 1,000 seconds. However, frequent
viewing may cause degenerative changes in the eye.
Class 11. Laser equipment emitting radiation (usually low-power visible
lasers) that presents a hazard when viewed directly for periods of
time longer than 0.25 second.
Class IIIa. Laser equipment emitting radiation considered harmful with di-
 Biomedical Deuice Technology: Princ$les and Design

rect viewing, especially when viewed with optical instruments.

Class IIIb. Laser equipment emitting radiation considered hazardous with
direct exposure to the skin and eyes.
Class IV. Laser equipment emitting radiation considered hazardous to the
skin and eyes from direct and scattered radiation.
The 0.25-second time interval is the average human reflex time for eye
closure when a visible beam of light hits the eye. For high-power lasers, even
the beam reflected from a shiny surface or scattered from a dull surface can
cause injuries. In medicine, most lasers are Class 111 and Class IV. The heli-
um-neon (HeNe) lasers used in the aiming beam for invisible lasers are usu-
ally Class I1 lasers.
To prevent eye damage, all personnel inside the operating room during
laser surgery must wear appropriate protective eyewear (eyeglasses or gog-
gles). Protective eyewear must be certified for the type of laser and of a spe-
cific optical density. The lens of the protective eyewear must attenuate the
laser beam to an acceptably safe level. It must be free from scratch and shield
the wearer's eyes from all directions of the visual field.

Skin Protection
Skin bums (patient or operating room personnel) can occur from expo-
sure to direct or reflected laser energy. Overexposure to ultraviolet lasers
may create skin sensitivity. To reduce the power density of the reflected laser
beam, metallic instruments with a polished surface should not be used dur-
ing laser procedures. The area surrounding the surgical site should be cov-
ered with fire-retardant materials.

Laser Plume
The smoke or laser plume arising from vaporization and charring of tis-
sue may become airborne from the surgical site into the surrounding atmos-
phere. Sample analysis has revealed that it contains water, carbonized parti-
cles, DNA, and intact cells. The plume has a distinct odor and may include
toxic substances, viruses, and carcinogens and therefore should not be
inhaled. The plume can scatter and attenuate the laser beam and obscure the
surgical site. Removal of laser plume enhances the visibility of the target site
for the surgeon. Smoke evacuation from the laser site and face masks can
prevent personnel from inhaling this plume. Laser smoke evacuators are
high-efficiency vacuum machines with submicron filters (to remove bacteria
and viruses) and active charcoal filters (to remove odor).
 Medical Lasers

Fire Hazards
Since a high-energy laser beam is used in laser surgery, operating room
personnel should be aware of and prepared for fire hazards. Flammable prep
solution should not be used. Fire-resistant drapes and gowns should be used.
A basin of sterile water should be available at the sterile site. A halon fire
extinguisher must be available in the operating room. Oxygen concentration
in the room should be as low as possible. Instruments used near the surgical
site must be nonreflective and noninflammable.

Access Control
During a laser procedure, only properly trained personnel with proper
eye protection should be allowed to enter and stay in the operating room.
Others must be aware of the hazards. To maintain a safety zone, the laser
operating location must be enclosed with access control. See-through win-
dows should be covered to prevent the laser beam from passing outside the
operating room (except for COz laser as it is absorbed by $ass). Warning
signs should be posted on the doors outside the operating room when lasers
are being used. The wording and symbols on these signs should be specific
for the type of laser in use. An example of a laser warning sign is shown in
Figure 35-8. Walls and ceilings should have nonreflective surfaces.
Reflective surfaces (glass on windows, mirrors, X-ray view boxes, etc.)
should be covered with nonreflective materials to prevent reflection of the
laser beam.

ISIBLE LASER RADIATION

AVOID EYE OR SKIN EXPOSURE
TO DIRECT OR SCATTERED RADIATION

5 J MAX PULSE, 20 MAX PPS at 1064 nm

CLASS IV LASER PRODUCT

Figure 35-8. Laser Warning Sign.

 Biomedical Device Technology: Principles and Design

Laser Safety Program

The CAN/CSA 2386 and ANSI 2136.3 Standards recommend laser
safety programs to be set up in workplaces using Class IIIB or Class IV
lasers. The program should include the following components: administra-
tive (develop laser policy, establish laser safety committee, etc.), engineering
control (install and maintain exhaust ventilation, window covers, etc.), and
personal protection (provide eye protections, appropriate training, etc.).
The Standards also recommend the appointment of a laser safety officer
(LSO) whose duty is to ensure the safe use of lasers in the workplace. Duties
of the LSO include:
Determine laser classifications
Ensure that laser equipment is properly installed and maintained.
Limit access to laser areas.
Arrange training for workers in safe use of lasers.
Recommend and ensure appropriate personal protection such as eye-
wear and protective clothing

MAINTENANCE REQUIREMENTS
AND HANDLING PRECAUTIONS

In addition to knowledgeable and trained staff, the performance and

safety of lasers rely on an effective preventive maintenance program.
Performance inspection of medical lasers and their accessories should be car-
ried out periodically to ensure that they conform to the manufacturers' spec-
ifications as well as current performance and safety standards. Output char-
acteristics, including laser power output, pulsing sequence, and timing accu-
racy, are measured by calibrated laser power meters. The gas mixture of a
gas laser must be replaced or recharged after being used for a period of time.
Failure to replace the gas will result in reduced laser power output. The lens-
es and mirrors in the laser delivery system are fragile; they are easily
scratched and damaged. Shock and motion from rough handling and even
from normal use will cause misalignment of the optical path in the hand-
pieces and laser arms. A misalignment in the delivery system of a laser will
result in little or no laser output. Optical alignment of the system should be
performed according to manufacturers' procedures. The transverse electro-
magnetic mode (should be close to a TEMw) of the laser beam should be ver-
ified after every optical alignment. Dirt on the first surface mirror will even-
tually lead to heat damage of the mirror from absorbing the laser energy.
Glass or silicon optical fibers are brittle and therefore cannot be bent too
 Medical Lasers

much. Care must be taken to inspect the tips of bare fibers or the tips of con-
tact laser probes for signs of cracks and heat damage.
 Chapter 36

ENDOSCOPIC VIDEO SYSTEMS

OBJECTIVES

Describe clinical applications of endoscopic video systems.

Analyze the construction and function of rigid and flexible endoscopes.
Evaluate essential features of an endoscopic video system.
Differentiate between fiberscopes and videoscopes.
State the advantages and disadvantages between a three-chip and sin-
gle-chip color CCD camera in endoscopy.
List common types of light sources for surgical video illumination.

CHAPTER CONTENTS

1. Introduction
2. Applications
3. System Components
4. Endoscopes
5. Light Sources
6. Camera, Processor, and Display
7. Image Management System
8. Insufflators
9. Advanced Development
10. Common Problems
 Endoscopic Video Systems

INTRODUcrION

An endoscopic video system allows the physician to look inside the

patient's body by inserting a light pipe with viewing optics (an endoscope)
into the body through a natural lumen or a small surgical puncture. The first
endoscopic instrument was developed in 1876 by Maximilian Nitze in
Austria. Endoscopic inspection of the abdominal cavity was introduced in
1902 and has since been refined and become widely used in many diagnos-
tic and therapeutic procedures. Endoscopy such as laparoscopic cholecys-
tectomy (removal of gallbladder using a surgical video system) or
arthroscopy has replaced many open surgical procedures. Endoscopic pro-
cedures are less traumatic to patients, cause less discomfort, and enable
shorter recovery time. Surgical procedures using endoscopy instead of open
I
surgery are often called minimally invasive surgeries (MIS).

APPLICATIONS

Laparoscopy refers to the minimally invasive treatment and examination

of organs and tissues in the peritoneal cavity using an endoscope and other
special instruments. In a multipuncture laparoscopic procedure, a small inci-
sion is made to allow the insertion of a trocar or cannula. A viewing laparo-
scope is inserted through the cannula. Another incision is made for the inser-
tion of the surgical instrument. Procedures such as cholecystectomy and
appendectomy can be performed by viewing the surgical site and inserting
the surgical instrument through the laparoscope without opening the abdom-
inal cavity. An external light source is required to illuminate the surgical site.
An insufflator helps to maintain a pneumoperitoneum to enlarge the work-
ing space for the surgical instruments and increase the field of view of the
surgeon within the peritoneal cavity. Similar to laparoscopy, arthroscopy
allows the diagnosis and treatment of some joint injuries and diseases with-
out open arthrotomy. In gastrointestinal and bronchial endoscopy, flexible
endoscopes are inserted through the rectum, esophagus, or trachea to allow
the diagnosis and treatment of diseases in the gastrointestinal and respirato-
ry tracts.
 Biomedical Device Technology: Principles and Design

SYSTEM COMPONENTS

A typical endoscopic video system consists of the following components:

An endoscope
A light source
A video camera
An image processor
One or two video display monitors
An image management system
Depending on the procedure, some of the following instruments and
devices may be used in endoscopic procedures:
Trocars/cannulae
Gas insufflators-COz or N20
Air, water, and suction pumps
Laser, ESU, U/S, cutters, forceps, scissors, biopsy snares, etc.
The following sections describe the basic components of a typical system.

ENDOSCOPES

An endoscope or viewing scope is used by the surgeon to view anatom-

ical structures and perform therapy to the interior of the body. The diameter
of an endoscope varies from the 1.7-mm needle fetoscope to the 5-mm
arthroscope to the 22-mm colonoscope. The length of the endoscope must
be appropriate to reach the desired structure. Endoscopes can be rigid or
flexible.

Rigid Endoscope
Rigid scopes (Figure 36-1) are either hollow sheaths that allow straight
viewing (such as laryngoscopes) or a sheath with an eyepiece and lens sys-
tem that allows viewing in a variety of directions (such as cystoscopes). The
sheaths of most rigid scopes are made of stainless steel, although plastic-
sheathed scopes (mostly disposable) are available.
A laparoscope is an example of a rigid endoscope. A viewing laparo-
scope employs a series of rod lenses to convey high-resolution, wide field of
view images to the eyepiece. Objects seen through a laparoscope may be
magnified or reduced depending on the distance between the object and the
tip of the scope. Optical fibers surrounding the rod lenses transmit illumina-
tion to the object from an external light source connected to the laparoscope
 Endoscopic Video Systems

Sheath

Objective lens Ocular lens

Fiber-optic light c a l k
Figure 36-1. External and Cross-Sectional View of a Rigid Endoscope.

via a fiber-optic light cable. An insufflator is connected to the laparoscope or

the trocar via an air hose. Through an air channel in the shaft of the laparo-
scope (or the trocar), NzO or COs gas is injected into the abdominal cavity.
The external diameter of a typical viewing laparoscope ranges from 5 to 10
mm. In an operating laparoscope, the eyepiece is offset from the shaft (by a
set of prisms) to allow insertion of the instrument through a separate instru-
ment channel in the shaft. The external diameter of a typical operating
laparoscope ranges from 8 to 12 mm. During a procedure, the object can be
viewed directly through the eyepiece. In practice, the eyepiece is often cou-
pled to a video camera and the images are displayed on a video monitor.

Flexible Endoscope
Instead of a rigid shaft, a flexible fiberscope has a long flexible insertion
tube connected to a proximal housing (Figure 36-2). Flexible endoscopes
can be inserted into curved orifices of organs such as colon, lung, and stom-
ach. To facilitate scope insertion and viewing, wires running from the control
head to the distal tip enable the user to angulate the distal end of the inser-
tion tube.
A flexible endoscope consists of the following sections:
Insertion tube
Control head
 Biomedical Device Technology: Princqles and Design

TIP angulat~on Optlcal or

Eyep~ecesect~on

cord

Figure 36-2. Flexible Endoscope.

Light guide connector

Universal cord (or light guide tube)
In a typical fiberscope, the insertion tube contains two bundles of optical
fibers, one for the illumination and the other for the image. A water channel,
an air channel, and an instrument channel are also included in the insertion
tube. Figure 36-3 shows the viewing and illumination optical pathway.
Figure 36-4 shows the cross-sectional view of the insertion tube. In a video-
scope, the viewing optical fiber bundle is replaced by a video camera chip
mounted at the tip of the insertion tube to pick up the image and convert it
to electrical signal directly. A videoscope generally has a larger diameter
than a fiberscope due to the larger size of the camera chip. To facilitate inser-
tion and viewing, the distal tip can be bent or angulated by moving a control
mechanism on the control head. The external diameter of the insertion tube
of a bronchoscope ranges from 0.5 to 6.4 mm with a working length from
550 to 600 mm, whereas it can be up to 14 mm in diameter and 2,000 mm
working length for a colonoscope.
During an endoscopic procedure, the physician holds the control head to
manipulate the insertion tube, introducing water or air to flush the site. The
control head houses the up/down and left/right angulation control knobs to
move the distal tip of the insertion tube as well as the aidwater and suction
control valves. The opening of the instrument channel is also located on the
control head.
The light guide tube is a flexible tube containing the fiber-optic bundle
for the light source. It also has separate air, water, suction, and COz channels
 Endoscopic Kdeo Systems

Ocular eyepiece Objective lens

Image fiber bundle

Figure 36-3. Image and Illumination Optical Path.

Suction channc

Instrument channel

Image light guide

Airlwater channel
Figure 36-4. Cross-Sectional View of Insertion Tube.

connected to those in the insertion tube via valves on the control head. The
light guide connector houses the adaptor for the fiber-optic bundle to the
light source. The connectors for air, water, suction, and COz (as well as the
electrical connector for videoscope) are also located on the light guide con-
nector. Figure 36-5 shows the water, air, suction, and COz channels of a typ-
ical gastrointestinal endoscope.
 Biomedical Deuice Technology: Principles and Design

MlllWQLPI VQIVP
Biopsylsuction channel
Suction valve \ Channel opening
Water channel

Universal cord
' I'
Insertion tube

-
Vacuum pump
air
water
suction

Suction bottle Air pump Water bottle

Figure 36-5. Air, Water, Suction, and Con Channels of an Endoscope.

LIGHT SOURCES

A light source is connected to the illumination light guide of the rigid or

flexible endoscope to provide illumination for viewing the surgical fields or
body cavities. Light sources are intended to provide the physician a sufficient
level of visible light for diagnostic observations and surgical procedures. A
light source usually emits a wide spectrum covering the visible, infrared, and
sometimes ultraviolet radiation. Infrared filters are installed in the light
source to prevent infrared radiation from entering the body, which otherwise
can cause thermal burn or even fire. A surgical light source can use a variety
of lamps, including xenon, quartz halogen, metal halide, and mercury vapor.
Xenon (color temperature from 5,600 to 6,600 K) and quartz halogen (from
3,200 to 5,500 K) are popular lamps for endoscopic procedures due to their
high intensity and near-daylight spectrum (5,000 to 6,000 K). The output
intensity of a light source can be adjusted either by an adjustable aperture or
by changing the brightness of the lamp. Changing the brightness of the
source by changing its supply voltage or current may be more energy effi-
cient, but doing so may alter the color temperature of the light. In systems
that have automatic brightness control, the light source is connected to the
video processor to automatically maintain the level of illumination through-
out the procedure. The output intensity and color temperature of light
 Endoscopic Edeo Systems

sources usually decrease with time. A typical xenon lamp has an approxi-
mate useful life span of 500 operating hours. Most lamps require forced cool-
ing to maintain a safe operating temperature.

CAMERA, PROCESSOR AND DISPLAY

With traditional rigid endoscopes and fiberscopes, an endoscopic camera

head is attached to the eyepiece (through an adapter) of the rigid scope or
the proximal end of the flexible endoscope. The most commonly used sin-
gle-chip mosaic color filter CCD (charge-coupled device) camera consists of
a single CCD chip with red, green, and blue colored filters overlaying each
CCD pixel. The light reflected from the object is filtered by each of the color
filters and incident on the underlying CCD elements. Each group of red,
green, and blue filters and CCD element combination forms one color image
pixel. The intensity of this filtered light is measured. After each exposure, the
mosaic signal from the CCD pixels is sent to the image processor. The three
signals (RGB) from each group of pixel are combined and reverted to the
color and intensity of the original light source. Another single-chip design
uses a rotating color wheel containing the red, blue, and green filters. In
essence, each CCD element is time-shared by the filters to measure the inten-
sity of the color components of the incident light. In a three-CCD-chip sys-
tem, the incoming light is split into red, green, and blue beams and each
beam is directed to one of the three dedicated CCDs. A three-CCD-chip sys-
tem provides a higher resolution image than the mosaic filter system and a
higher refreshing rate than the rotating color wheel system.
In a videoscope, the CCD chip is integrated into the tip of the scope to
provide a high-quality picture free from image distortion and degradation
from optical misalignment and deterioration of the optical fibers and lens
system. However, this may increase the size or diameter of the endoscope.
The image or video processor is responsible for white balance, bright-
ness, contrast, color control, focus, and shutter control. Some video proces-
sors can support automatic gain control (AGC), multiformatting, and char-
acter generation.
During a minimally invasive surgical procedure, video or still images are
displayed on one or more color monitors. Typical medical-grade video mon-
itors have low leakage current with high brightness and high contrast, allow
gamma curve calibration, and can support resolution better than 800 lines
per frame. Both cathode ray tube (CRT) monitors and liquid crystal displays
(LCD) are used in endoscopic systems.
 Endoscopic Video System

COMMON PROBLEMS

Two major complications of endoscopy are:

1. Perforation-Perforation is a major cause of concern when rigid
scopes are used. The risk of perforation for flexible scopes is lower
but it remains a potential complication.
2. Bleeding-Excessive bleeding can occur from a biopsy site, removal
of a polyp, or areas where tissue has been cut.
Patients may also suffer from burn injuries when electrosurgery is used
with endoscopy.
Apart from mechanical wear and tear and electronic component failures,
problems associated with endoscopic system can be separated into three cat-
egories:
1. Cleaning, disinfection, and sterilization
2. Water leakage
3. Optics
All parts of endoscopes and accessories must be thoroughly cleaned, dis-
infected, or sterilized after every use. The primary causes of infection trans-
mission in endoscopic procedures are inadequate cleaning and disinfection
or sterilization of the endoscopic equipment. It is important to clean all parts
of the endoscope as soon as possible after use while organic debris is still
moist. Tissue, mucus, blood, feces, and protein residue can become trapped
in the channels and are difficult to remove when dried. Nonresidual liquid
detergent are used with small brushes to clean the inside lumen of all chan-
nels. Special ultrasonic cleaners may be used. If the scope can be sterilized
with ethylene oxide, it must be thoroughly cleaned and dried before steril-
ization. Scopes that cannot be sterilized are disinfected by soaking them in
activated glutaraldehyde, acetic acid, or formaldehyde solution. Care must
be taken to fill all the lumens with the disinfectant. Endoscope processors are
available to clean and disinfect flexible scopes. An automatic endoscope
processor cycles through wash, disinfect, rinse, and dry to clean and disinfect
the scope once it is set up properly in the washer.
An insertion tube of a flexible endoscope is covered with a waterproof
plastic sheath. If this waterproof sheath has a leak, water or bodily fluid will
enter the internal part of the scope. Water leak will cause deterioration of
internal components (e.g., rusting of the angulation cables) and prevent effec-
tive disinfection or sterilization. Moisture will fog up and damage optical
components. Leak tests on flexible endoscopes should be done on a regular
basis.
Optical fibers in the light guide or image bundles may break, causing
lower light illumination or black spots on the image. Poor handling of the
 Biomedical Device Technology: Principles and Design

scopes may damage the lens or cause misalignment. Moisture inside the
sheath may decrease transmission and damage the optical components.
The heat from the intense light source may cause patient bum. Although
infrared filters in the light source are used to remove I R radiation, care must
be taken not to shine the light onto the same position for a prolonged peri-
od of time. Radio frequency (RF)leakage current may cause a secondary site
bum in a patient undergoing an endoscopic electrosurgical procedure.
Electrical leakage and insulation tests are performed periodically on endo-
scopes to detect potential current leakage problems.
As most injuries are internal, they may not be observable immediately
after the procedure. Moreover, extreme care must be taken to observe the
patient after an endoscopic procedure. Ambulatory outpatients must not
leave the facility until vital signs are stable and the surgical side effects have
passed.
 Appendix A-1

A PRIMER ON FOURIER ANALYSIS

Consider a symmetrical square waveform with amplitude o f f 1 V and a

period of 1 sec. (Figure 1)

Amplitude (V)
A

b
0
0.5 1.O 1.5 Time(s)

-1

Figure 1. Square Waveform in Time Domain.

This signal can be expressed by the following summation of sinusoidal

signals, known as the Fourier series:
V(t) = 4V/n (sin 2nt + 0.33sin6nt + 0.20sinlOnt + 0.14sinl4nt + . . .).
The sinusoidal components in the Fourier series constitute the frequency
spectrum of the square wave signal. Such a spectrum can be graphically rep-
resented in Figure 2, where the horizontal axis represents the frequency in
Hz. The lowest nonzero frequency fo of the spectrum, in this case equal to 1
Hz, is called the fundamental frequency and the others, which are in multi-
 Biomedical Device Technology: Princ$les and Design

ples of fo, are called the harmonics. In the case of this square wave signal, the
frequency spectrum contains the fundamental frequency and only the odd
harmonics (i.e., 3rd, 5th, 7th, etc.).

Amplitude (V)

A = 41.T

I 1 3 5 7 9 Frequency (Hz)

Figure 2. Square Wave (Fig. 1) in Frequency Domain.

Note that the amplitude of the higher harmonics decreases with increas-
ing frequency. In general, harmonics of very low amplitude are insignificant
and can be ignored (which means that the signal is considered to have a finite
bandwidth).
Figure 3a shows three sinusoidal waveforms of frequencies lHz, 3Hz,
and 5Hz with amplitudes equal to 1.0, 0.333, and 0.20 V, respectively. The
three waveforms can be represented in mathematical form as:

Figure 3b shows the sum of the waveforms Vi + V2 and Figure 3c shows

+
V1+ V2 V 3 .
We note that as more waveforms are added together, the more it will
resemble a square wave. As it turns out, if we keep adding the higher har-
monics, we will eventually get back a perfect square wave like we have seen
in Figure 1. This simple example illustrates that a signal in the time domain
can be fully represented by its frequency and phase spectrum in the fre-
quency domain (note that we have not discussed the effect of phase shifts of
the harmonics).
We have shown that the frequency spectrum for a periodic square wave
is composed of discrete frequency components. These frequency compo-
nents are at multiple frequencies of the fundamental frequency. Other than
 A Primer on Fourier Analysis

Amplitude (V)

Amplitude (V)

(b)V = Vl + Vz

Arnpl~tude(V)

(c)v = v, + v2 + v3

Figure 3. Frequency Component of a Square Wave (Fig. 1).

 Biomedical Deuice Technology: Princ$les and Design

periodic signals, Fourier transform may be applied to a nonperiodic function

of time. A nonperiodic signal produces a continuous frequency spectrum
(instead of a discrete spectrum). That is, unlike the spectrum of a periodic
function of time, the frequency spectrum of a nonperiodic waveform spreads
over the entire bandwidth, as shown in Figure 4.

Amplitude (V)

Figure 4. Frequency Spectrum of a Nonperiodlc Waveform.

Figure 5 shows a time domain signal of a blood pressure waveform and

its frequency spectrum. (Note that the signal has no significant frequency
components above 9 Hz and it has a positive amplitude at zero frequency
due to a nonzero average pressure in the time domain.)

A
A

b I b
Time (sec) 10 Frequency (Hz)
(a) (b)
Figure 5. Blood Pressure Waveform in (a) Time and (b) Frequency Domains.
 Appendix A-2

OVERVIEW OF MEDICAL
TELEMETRY DEVELOPMENT

1. INTRODUCTION

Purpose
Medical telemetry is defined (by the AHA'S Spectrum Selection
Workgroup) as
The wireless transfer of information associated with the measure-
ment, control, and/or recording of physiological parameters and
other patient related information between points separated by a dis-
tance, usually within the healthcare institution.
Note that the modulated signal can also be transmitted via hard wires
such as a telephone network. The most common patient vital sign transmit-
ted in patient monitoring systems is the ECG waveform. Telemetry has sub-
stantially reduced the risk to patients who may otherwise require continuous
monitoring.
Typically, a telemetry ECG transmitter would be used on a patient who
has been released from ICU and is now in a step-down inpatient ward. It is
critical for the patient's recovery that they become ambulatory. In the past,
without telemetry, this was not possible due to the monitoring requirements.

Aduantages in use of telemetry

Mobility of subject under study or monitor
Minimal disturbance of routine
Centralization of expensive equipment (e.g., automatic arrhythmia
monitors)
 Biomedical Device Technology:Princ$les and Design

Examples of biomedical applications

ECG, EEG, Temp., Resp., EMG, pH transmission
Position localization
Stimulation via telemetry
Pacemaker programming and information download
Transmission of image information over long distances (e.g., teleradi-
ography)

2. PROBLEMS WITH TELEMETRY

Problems with telemetry are primarily related to data rate and reliabili-
ty. Some factors are:
Bandwidth requirements
Channel overcrowding
EM interference and immunity
Transmission range
Power requirement for mobile units (MU)
CRTC/FCC licensing
Primary (registered users who have the right to use the bandwidth)
versus secondary (users who do not have the exclusive right to use the
bandwidth) users

3. DEVELOPMENT AND TREND IN MEDICAL TELEMETRY

Development
Started in the VHF band (174-216 MHz) Using Analog Modulation
Techniques
Migrate to U H F (460-470 MHz) using digital modulation.
New wireless medical telemetry system (WMTS) bands: 608-614
MHz in 2000, 1,395-1,400 MHz in 2006.
The 2.4 GHz Industrial, Scientific and Medical (ISM) band limits
transmission power and requires users to use spread-spectrum tech-
nology. Users are all sharing the bandwidth with equal rights.
Currently, some manufacturers use the ISM bands; others choose to
use the WMTS band.
Medical Telemetry Using General VHF Band
 Overview of Medical Telemetry Development

174-216 MHz
Nonprimary user
Sharing frequencies with other nonmedical users (e.g., TV Channels
7- 13)
Unidirectional
No voice or video
Obsolete
Medical Telemetry Using General UHF Band
460-470 MHz
Nonprimary user
Sharing frequencies with other no-medical users
Unidirectional
No voice or video
Will be phased out
Medical Telemetry using WMTS Band
608-614 MHz, 6 MHz bandwidth, and 1,395-1,400 MHz (will
expand to include 1,429-1,432 MHz)
Dedicated band for medical telemetry
Unidirectional
No voice or video
Primary user protected against intentional interference but not out-of-
band interference (e-g.,EM1 sources from other medical or nonmed-
ical devices such as foot massagers)
Medical Telemetry Using ISM Band
2.4000-2.4835 GHz, 83.5 MHz bandwidth
Allow voice and video data (e.g., can use VoIP)
All users must use spread-spectrum technology
Wireless Ethernet: IEEE802.11 (a wireless extension of Ethernet:
IEEE802.3)
802.11: 2.4 GHz, 1 and 2 Mb/s using DSSS or FHSS
802.11b: 2.4 GHz, 11 Mb/s using DSSS
802.11a: 5.7 GHz, 54 Mb/s using OFDMSS
WLAN connects to LAN via an access point (AP)

4. SPREAD SPE(;TRUM TECHNOLOGY

The characteristics of spread-spectrum technology are:

Use packet (short burst of data) transmission
 Biomedical Device Technology: Princ$les and Design

Bandwidth of transmitted signal is spread over a much greater band-

width than the original signal
79 channels available in North America (79 X 1 MHz wide channels
from 2.402 to 2.480 GHz)
A two-step modulation process-one modulation step to spread the
data and the second step to modulate the spread signal (e.g., the sec-
ond step uses frequency modulation)
Increased processing gain
Robust-high resistance to noise and interference
Bidirectional-allow data retransmission in case of interference
(receive acknowledge). May reduce throughput but ensure data
integrity.
Allow two-way communications
Low spectral power density: average energy in a specific frequency
band is very low; less chance of signals interfering with other systems
Three methods: direct sequence spread spectrum (DSSS), frequency
hopping spread spectrum (FHSS),orthogonal frequency division mul-
tiplexing spread spectrum (OFDMSS)
 Appendix A-3

MEDICAL GAS SUPPLY SYSTEMS

In a typical acute care hospital, medical gases are available through

piped-in wall outlets. A typical operating room is equipped with oxygen,
nitrous oxide, medical air, nitrogen gas, and suction outlets on the wall or on
the ceiling column. At the bedside of a typical patient ward, oxygen, medical
air, and suction are available. Cylinder gases are used to supply some less
common gases and also used as backup in case the central supply is inter-
rupted. The pressure for piped-in gas supply is usually about 50 to 55 psig.
The absolute pressure of wall suction is usually about 200 mmHg.
Figure 1 shows a typical cryogenic bulk central supply system. During
normal operation, gas is drawn from the primary operating supply reservoir.
A secondary operating supply reservoir is used when the primary is deplet-
ed. The primary reservoir is filled soon after the supply has been switched to
the secondary reservoir. To ensure uninterrupted supply, a number of gas
cylinders are connected to the central supply line as reserve supply. These
cylinders are checked regularly but will not normally be used to supply the
system.
Four sizes of gas cylinders are available. The dimensions are tabulated in
Table 1.

Table 1.
Gas-Cylinder Dimensions
- -- - -- -

Qlinder S i a Height w/Cap (in.) Outside Diameter (in.)

- - - -- - -. - -
D 20 4
 Biomedical Device Ethnology: Principles and Desip

-
To 50 psig gas
piping system

regulatingvalve

manual valve
closed

manual valve
open
Primary Secondary Reserve supply
operating operating
supply supply

Figure 1. Medical Gas Central Supply System.

Table 2 shows the characteristics of common medical gas cylinders. The

pressure at room temperature, capacity (H-cylinder),color coding, and the
state of the gas inside the cylinder of the gases are listed.
In order to prevent a wrong medical gas cylinder from being connected
to the gas line or the inlet of a device, a pin-indexed safety system (PISS) is
used. In this safety system, the stem of a gas cylinder can be allowed to con-
nect only to the cylinder yoke of the same gas. A pin-indexed safety system

Table 2.
- - - - -
Cylinder
-
Data of Common Medical Gases
-- - - - - - - - - - - - - -

Gas
- - -
Pressure
- -
(psig) Capacily (liters)
- - -- - - -
State
- - -
Color Code
Oxygen 2,217 7,000 Gas White
Nitrous oxide 745 15,540 Liquid Blue
Medical air 2,217 6,500 Gas Black and white
Helium 2,217 8,200 Gas Brown
Carbon dioxide 838 12,360 Liquid Gray
Nitrogen 2,2 17 6,400 Gas Black
- - - - - - - - - - - - - - - -
 Medical Gas Supply System

uses a set of pins on the yoke and a set of holes on the stem to encode the
medical gases. The cylinder can connect to the yoke only when the pins are
at the same matching location as the holes. For the same gas, the location of
the pins on the yoke aligns with the locations of the holes on the stem.
A diameter-indexed safety system (DISS) is designed to prevent a wrong
hose from being connected to the piped-in outlets. In this system, the diam-
eter of the connector on the flexible hose is encoded together with the con-
nector of the wall outlet for the medical gas. Only the hose connector of the
gas can be connected to the piped-in wall gas outlet of the same gas. The flex-
ible hoses are color-coded according to the gases to further minimize con-
nection errors.
 INDEX

o/oSa02,470-472 Adaptive filtering, 477

'%lsp02,470-471 ADC (see analog to digital conversion)
o/oSv02,470 Adolf Fick, 324
10-20 system, 267, 270 AGC (see automatic gain control)
left auricular point of, 270 AHA (see American Heart Association), 316
inion of, 270 Air temperature, 450
nasion of, 270 Airway resistance, 411
right auricular point of, 270 AK (see artificial kidney)
12-lead ECG, 247,249 Albert Einstein, 521
Alkaline cell, 146
Alveolar gas exchange, 423
Alveolar pressure, 410
AAMI BP22/BP23 Standards, Alveoli, 410
305 Ambulatory ECG, 244
Ablation, 392 Ambulatory monitor, 229
Abnormal fetal heart rate American Heart Association, 316, 551
bradycardia in, 444 American Society of Anesthesiologists, 470
tachycardia in, 444 AMLCD (see active matrix liquid crystal dis-
variation in, 444 play)
Abrasion, 526 Amperometry, 142
Absolute pressure, 55 Amplification, 40
Absolute temperature scale, 67 Amplification factor, 40
Absorbance, 471 Amplitude linearity, 46
Absorptivity, 471, 473 Amplitude-Zone Time-Epoch-Coding, 232
Acceleration, 88 Analgesic, 487
Access point, 553 Analog modulation, 552
Accuracy, 30 Analog to digital conversion, 26
Acetic acid, 545 Anatomic dead space, 411-412
Action potential, 150, 165 Anesthesia, depth of, 265,
membrane, 11-12 272-273, 497-498
single cell, 12 Anesthesia machine, 486-498
Activated carbon, 518 adjustable pressure-limiting valve of,
Active breathing, 410 494-495,497-498
Active cancellation, 175 agent monitor of, 497-498
Active electrode, 392, 394 bellow of, 494
Active matrix liquid crystal display, 214 breathing and ventilation subsystem of,
Actuator, 43 487,492-495
 Biomedical Device Technology

breathing bag of, 493 template cross-correlation, 231

check valve of, 488 template matching, 231-232
C 0 2 absorber of, 493-494,497-498 waveform feature extraction, 231
C 0 2 absorption canister of, 493 Arrhythmia, 224, 338, 356
color-coded gas cylinder of, 497-498 definition of, 338
continuous flow and rebreathing, 487 Arterial blood, 325
flow control valve of, 488-489 Arterial blood pressure monitor
flowmeter of, 488 amplifier of, 306
functional block diagram of, 488 bandwidth of, 306
gas cylinder of, 489-490 catheter error in, 308
gas supply and control subsystem of, common problems of, 308
487-491 display of, 307
hanger yoke of, 489-490 filter of, 306
injury related to, 496-498 functional block diagram of, 304
manual breathing mode of, 492-493 setup error in, 308
oxygen flush valve of, 489 signal isolation of, 306
oxygen ratio monitor of, 496, 498 signal processing of, 307
pop-off valve of, 493 spectral analysis of, 306
pressure gauge of, 490 transducer of, 303-306
pressure-limiting valve of, 494-495, zero offset of, 306
497-498 Arterial blood pressure monitoring
regulator of, 488, 490 blood clot in, 300
scavenging subsystem of, 487,493, catheter of, 300
495-496 continuous flush valve of, 300
shutoff valve of, 488-489 extension tube of, 300
touch-coded control knob of, 496, 498 functional block diagram of, 303-304
vaporizer interlock of, 497-498 heparinized saline in, 300
vaporizer of, 489-490 offset in, 300, 302
ventilator mode of, 492, 494 patient port of, 300, 302
Anesthetic agent, 487-490 pressure transducer of, 300
Anesthetic agent, occupational hazard from, rapid flush valve of, 300
495 setup of, 300
Anesthetic gas, 487-488, 492 transducer port of, 300,308
Angular motion transducer, 88 zero port of, 303
Anions, 128 zeroing process in, 300,
Anode, 128, 145 302-303
ANSI Z136.3-1988 Standards, 531 Arterial blood sample, 470
ANSIIAAMI HF18 Standards, 399 Arterial hypoxemia, 410
Antepartum monitoring, 444 Arterial line, 300
Antibradycardia pacing, 358 Arterial oxygen saturation, 470
Antitachycardia pacing, 358 Arterial tonometry, 321-322
APL valve (see adjustable pressure-limiting Arthroscopy, 537-538
valve of anesthesia machine) Artificial kidney, 500, 503
Apnea, 418,429 blood compartment of, 500
Appendectomy, 537 dialysate compartment of, 500
ARCnet network protocol, 235 semipermeable membrane of, 500
Argon laser, 525 Atmospheric pressure, 54
Argon-enhanced ESU, 392 ATP (see standard atmospheric pressure)
Arrhythmia detection, 231-233 Atrial fibrillation, 356, 368
Arrhythmia detection algorithm, Atrial flutter, 356-358, 368
 Index

Atriwenhidar node, 24Q,$46 46%

Aium,240 Bipolar electrode, 3 5
~%&~~PEO 40Q Bipolar lead,341
& p n b d limb h d I24?%4& EIS [sa lbispectral index)
Au~~;zzig;tBzyg l ~Bl2
~, &i~"pedtralin&x, 265,487,498
hbm& gab wntr~1,%3 Bi8pt:cttal index m@&hring,4974%
Auterrnatk NIBP modw, 517 Bimntricular pacing, 342
9s,Y+&?&hy Skl Bhrc;k*, Elk-I&%, 464
A-V @aft, 51 1 Bladder pmsure, 295
&V node (5cb-MvmWmls~r model, 8M Biended mode, 395
A-V shunta528 Blodk.d-s, i&
Axial strain, 59 Blood mlls, hamolysis of, 372
Axonal velocity, 290 Blood clot, 514
aaEC (seeAmpBtude-Zone Time-Epa& Blood flow, 14
coding) Blood flowmetex, 435
Bled gw,477
P RdwW*\@
Blood pH, 14,410
B a ~ t e r kwlture, 518 Blood pressure, 14 295-322,448
Balanced bridgeJ63 2xctmi& 2%'-2%
Hand @p energy, 114 venom, 297-298
Bandwidth, M, IS$, 235+244%284 272,306, wntzti~ulat,297
5a,59 B1wd p w m e &&er mm
*WIe&rn

fl

reeardee af,2% cemd flmting blcudk of, 383

tredfng cap&ility of, 228 diaphragm d,304
Benow, .57-5$3 excfWon ofa304
Bernoulli's ~ q W o n97,
, 100 ~~ezore~stive $traingauge of?304
Bili-rubin, 450,453 peeare dome of, 304
B2nekd'Eicmmmr, 69 resistive strain gauge of, 3303
Biommp;stibili~~ 22 sensitivity of, 304-905
Biam&al Ggaal measuremexlts coiaSMnIY6: ,++&minuYiw oif, 304
in, %a f k d y temperature manitass,457-468
Bjopakntlal, 16, 14&-1&9,159 Blood compatibi&b,r,%3
o w ~ f 1% * M9 Bramirag bdl a d h c e p e , 21 I
Biqmtmtial mp&krs, 15&M1 Baurdaa tabI5657
Et'i&enWek&de%,1483-152 BPW*S45
h p o t a & d signal, 165 Bran d&, 265
Bmxrurrsrs, 144 Brain &terny264
Biphmic truwa@dexponentid wav~fom~ IE;leakdow734
 Biomedical Device Technology

Bridge, network, 237 inverter of, 366

Bronchoscope, 540 monophasic waveform of, 358-359
BTPS, 416, 428, 433 output isolation of, 366
Bundle branches, 240 patient load of, 361
Bundle of His, 240, 338 power supply of, 360
Buried channel MOS capacitor, 119 quality assurance of, 368
voltage monitor of, 365
wave shaping inductor of, 360
waveform shaping circuit of, 360, 365
Cardiac output, 338, 340 Cardiac index, 324
Cardiac output monitor, functional block dia- Cardiac output, 14, 324, 356, 502
gram of, 333-334 Cardiac output monitor, 323-336
Calibration, 33 Cardiac output monitoring
Calibration factor, 310-31 1 diffusible indicator of, 327
Calomel reference electrode, 136-137 indicator of, 326-327
CAN/CSA-2386.92 Standards, 531 tracer of, 326
Cannula, 510,537-538 Cardiac vector, 242
Capacitive coupling, 167 Cardiopulmonary system, 470
CAPD (see continuous ambulatory peritoneal Cardiovascular system, 298-299
dialysis) Cardioversion, 356-357
Capnogram, 481 synchronization circuit of, 368
Capnograph (see also end-tidal C 0 2 moni- Carotid artery, 465
tor), 482 Carotid artery occlusion, 435
Carboxyhemoglobin, 471-472 Cartridge filter, 518
Cardiac activity, 240 Cathode, 128, 145
Cardiac arrest, 356 Cathode ray tube, 209-212, 307, 543
Cardiac arrhythmia, QT interval of, 242 anode of, 209
Cardiac care unit, 224 brightness of, 217
Cardiac cycle, 297-298, 307 cathode of, 209
Cardiac defibrillation, 340 contrast ratio of, 212, 217
Cardiac defibrillator deflection plate of, 209
biphasic waveform of, electron beam of, 209
358-359,362-364 electron gun of, 209
charge control, 360, 365 phosphor screen of, 209
charge relay of, 361, 365 refreshing rate of, 212, 217
charging circuit of, 360 resolution of, 212, 217
common problems of, 369 triggering of, 210
contactor of, 364 viewing angle, 217
current monitor of, 365 Cations, 128
defibrillator paddles of, 365 Cauterization (see also coagulation), 392, 401
discharge buttons of, 365 CCD (see charge-coupled device)
discharge control of, 360 CCD Array, 120, 122
discharge relay of, 361, 365 CCD Pixel, 119
energy delivered in, 361-362 CCPD (see continuous cycler-assisted peri-
energy dumping in, 362, 366 toneal dialysis)
energy storage capacitor of, 360 CCU (see cardiac care unit)
energy stored in, 361 Celsius temperature scale, 66
functional block diagram of, 360, Cell diagram, 129
365-366 Cell membrane, 10
inrush current of, 361 Cellulose acetate, 508
 Index

Central chart recorder (see central recorder) 517

Central nervous system, 263-264 Continuous cycler-assisted peritoneal dialysis
Central recorder, 226,230 dispasd bag of, 516
Central station, 229 drain bag of, 516
basic capabilities of, 229 heater compartment of, 516
Cerebellum, 264 supply reservoir of, 516
Cerebral cortex, ridges and valleys of, 264 volume control compartment of, 516
Cerebral spinal fluid, 264,266 Continuous laser, 525
Cerebrum, 264 Continuous mandatory ventilation, 426
Charge-coupled device, 118- 124 Continuous paper feed recorder
stop region of, 119 building blocks of, 206
trapped charge of, 119 Continuous paper feed recoder, 204
Chemical disinfection, 518 Continuous positive-airway pressure, 426
Chest lead, 247-248 Continuous temperature mewrernent, 45$
Cholecystectomy, 537 Continuous temperature monitors, 459
Cholesterol esterase, 144 Control inputs, 17
Cholesterol oxidase, 144 Conmned mandatory ventilation, 426
Clark polarograhic oxygen electrode; 142 Convection, 458
Classical instrumentation amplifier, 163 Conversion factors, 54
Clearance, 503,507,509, 515 Co-oximeter, 470, 474
plasma, 503,507 Core tlemperature, 459,462i2,465
CMG (see common mode gain), 161 Corner frequency (see cutoff frequency)
CMRdB, 165 Coronary care unit (see also cardiac care
CMRR (see common mode rejection ratio) unit), 230
CMV (see continuous mandatory ventilation) Cortex, 238,264-265
CNS (see central nervous system) premotor, 264
C Q (see cardiac output) primary motor, 264
C 0 2 monitor, 481 CPAP (see continuow positive-air-way pres-
Coagulant, 514 sure)
Coagulation, 392,401 Crawfasd Long, 486
eioagulation mode, 395 Critical care ventilator, 424
Cognitive limitation, 15 Cross-talk, 353
Coherent, 521 CRT (see cathode ray tube)
Collimated, 521 Cryogenic bulk central supply sptem, 555
Collodion, 267 CSF (see cerebral spinal fluid)
Colonoscope, 538,540 GSMklCD, 235
Color temperature, 112-1 13 cuprophane, 568
Common aorta, 296-297,299,325 Current density, 392
Common mode gain, 161 Cut, 392,394
Common mode rejection ratio, 161,292,306 Cut inode, 395
Common mode signal, 163 Cutoff frequency, 39
Compressed gas cylinder, 429 CV (see controlled mandatory ventilation)
Conduction, 458 Cystascope, 538
Conduction disorder, 339
Conduction velocity, 290-291
Conductive interference, 179-181
Conductivity meter, 512 DAC (see digital to analog conversion)
Conductor loop, 177 Damped sinusaidal waveform, 394
Contact temperature sensor, 458 Daniel Bernoulli, 97
Continuous ambulatory peritoneal dialysis, D a ~ e lcell,
l 129
 Biomedical Device techno lo^
Dark current, 116, 123 379-380
Data management, 487 Diastolic blood pressure, 296, 298-299, 307,
Data transfer rate, 236 314-317, 320, 322
Datapoint, 235 DICOM (see Digital Imaging and Commu-
DC defibrillator, 358 nications in Medicine)
DC offset, 155 Dicrotic notch, 298-299
Dead space air, 414,481 Differential amplifier, 154, 160,292
Dead zone, 34 Differential gain, 161
Defibrillator protection, 180-181 Differential mode signal, 163
Defibrillator/monitor, 356 Differential pressure flow transducer, 415
Deoxygenated blood, 296,325 Diffusion, 149, 501, 505, 511
Deoxyhemoglobin, 470-473 Diffusion force, 10
Depolarization, 240 Digital Imaging and Communications in
membrane, 11 Medicine, 238
Depolarized cell, 150 Digital Processing, 26
Desflurane, 49 1 Digital to analog conversion, 26
Desiccation, 392-394 Direct Fick method, 325
Desired signal (see also desired input), 166 Direct input, 45
Dextrose, 328, 517 Direct sequence spread spectrum, 554
DG (see differential gain) Displacement, 88
Diagnostic device, 5,26 Displacement current, 168-169
Diagnostic ECG, 243-244 Displacement transducer
bandwidth of, 244 capacitive, 9 1-92
Dial-type temperature gauge, 68 inductive, 89-91
Dialysate delivery system resistive, 88-89
single-pass recirculation, 512 Display system
single-pass, 5 12 first-order high pass filter of, 220
sorbent regenerative, 512 frequency response of, 218-222
total recirculation, 513 lower cutoff frequency of, 219
Dialysate, 504, 511-513 paper speed of, 218
acetate base, 511 performance characteristics of, 216-222
batch processing of, 511 resolution of, 218
bicarbonate base, 511 sensitivity of, 218
composition of, 512 step response of, 220-222
Dialysis equipment, 499-519 transfer function of, 219
Dialyzer (see also artificial kidney), 506-510 upper cutoff frequency of, 219
clotting properties of, 510 Display technologies, comparison of, 217
coil, 508 Disposable transducer, 303
high flux, 507 blood pressure, 305
high permeability, 507 DISS (see diameter-indexed safety system)
hollow fiber, 508-510 Dissipation constant, 69, 75
parallel plate, 508 Distillation, 518
priming volume of, 509 Doppler blood flowmeter, 439-442
reuse of, 519 audio frequency amplifier of, 441
transmembrane pressure of, 507,515 audio speaker of, 441
ultrafiltraion coefficient of, 507-509 blood vessel-wall motion in, 441
Diameter-indexed safety system, 497-498, demodulator of, 441
557 functional block diagram of, 441
Diaphragm, 410 integrator of, 442
Diaphragm pumping mechanism, 373, RF oscillator of, 441
 ultrasound receiver of, 441 invasive electrode of, 267
ultrasound transmitter of, 441 nasopharyngeal electrode of, 267
zero crossing detector of, 442 needle electrode of, 267
Doppler effect, 437, 439 subdural electrode of, 267-268
Doppler flowmeter (see also Doppler blood surface electrode of, 267
flowmeter), 439 EEG electrode and placement, 266-272
Doppler shift, 436-437, 439-442 Effector limitation, 16
-

Doppler ultrasound~bloodpressure monitor, ' Einthoven's Mangle, 244

320-321 Electnc arc, 394
cuff pressure of, 320 Electrical hazard, 184
Doppler shift in, 320 reasons of increase of, 184
Double insulation, 198 summary of, 189
Double layer capacitance, 155 Electrical safety, 27, 183, 462
DSSS (see direct sequence spread spectrum) Electrical shock, 183-184
Dual chamber pacemaker, 342, 346 Electrical shock hazard, 183
Duty cycle, 395 Electrical stimulation, 338
Electrocardiogram, 13, 150, 242-243, 245,
E 247,253,259
common problems of, 259-261
E. J. Marey, 313 definition of, 242
Ear thermometer (see also tympanic ther- P wave of, 242
mometer), 463 power frequency interference of, 259
ECG (see electrocardiograph and electrocar- PQ interval of, 242
diogram) QRS complex of, 242
ECG data management system specifications R wave of, 242,357,368
of, 257 surface, 242
ECG lead, 245,251 T wave of, 242,356-357,368
lead aVF of, 247 Electrocardiograph, 14, 168, 239-261
lead aVL of, 247 amplifier of, 256
lead aVR of, 246,251 calibration pulse of, 256
lead I of, 245,251 defibrillator protection of, 254
lead I1 of, 245 filter of, 256
lead I11 of, 246 frequency bandwidth of, 244, 256
lead V1 of, 247-248 functional block diagram of, 254
lead V2 of, 247-248 lead selector of, 255
lead V3 of, 247-248 lead-off detector of, 254
lead V4 of, 247-248 notch filter of, 257, 260
lead V5 of, 247-248 preamplifier of, 255
lead V6 of, 247-248 recorder or display of, 257
ECG lead configuration, 245-252 right-leg-driven circuit of, 256
ECG monitor, 352 signal isolation of, 256
Ectopic focus, 243 signal processor of, 257
Eddies, 96 Electrochemistry, 126
EEG (see electroencephalography or electro- Electroconductive pathway, 356
encephalogram) Electrocorticography, 263
EEG electrode Electrocution, 183
cortical electrode of, 267 Electrode double layer, 152-153, 156
depth electrode of, 267, 269 Electrode gel, 156
earlobe electrode of, 267 Electrode-electrolyte interface, 156
impedance of, 272 Electrodes, 43, 134-143
 Biomedical Device Technology

half-cell potential of, 152 Electron current, 151

nonpolarized, 152, 155 Electroneurophysiology
perfectly nonpolarized, 153 definition of, 263
perfectly nonreversible, 152 baseline wander of, 282
perfectly polarized, 152, 153 Electronic thermometer, 462
perfectly reversible, 153 Electrophysiology study, 152
polarized, 152, 155 Electrosurgery
Electroencephalogram, 263,265-266 active electrode of, 397
alpha wave of, 265,272-273 common problems of, 406-408
beta wave of, 265,272-273 fire and explosion in, 407
delta wave of, 266,272-273 modes of, 394
theta wave of, 272-273 muscle and nerve stimulation in, 408
Electroencephalograph, 14,262-280 return electrode of, 397-399
analog to digital converter of, 276 Electrosurgical unit, 391-408
artifacts in, 279 burst repetition frequency of, 401
bandwidth of, 272 capacitive leakage current of, 406
bipolar input connection of, 275 functional block diagram of, 401
chart speed of, 278 high frequency leakage test of, 406
electrode impedance tester of, 278 isolated output of, 402
errors and problems of, 279 output characteristics of, 403-404
filter of, 278 output power of, 403
functional block diagram of, 273-278 output power verification of, 405
head box, 274 patient load of, 403
sensitivity of, 278 percentage isolation of, 406
signal isolation of, 276 power amplifier of, 401
transverse bipolar, 275, 277 quality assurance of, 405-406
troubleshooting of, 280 step-up transformer of, 401
unipolar input connection of, 275 EMG (see electromyography or electromy
Electroencephalography, definition of, 263 gram), 263
Electroluminescent display, 216 EMG & EP studies
brightness of, 217 concentric needle electrode of, 285
contrast ratio, 217 end plate noise in, 287
refreshing rate of, 217 end plate spike in, 287
resolution of, 217 fasciculation in, 287
viewing angle, 217 filter of, 29 1
Electrolyte, 152 grounding electrode of, 283
Electrolyte gel, 368 insertion activity of, 286
Electrolytic cell, 126 monpolar electrode of, 285
Electromagnetic flowmeter, 104 motor response of, 288
Electromagnetic immunity, 408 needle electrode of, 284-285
Electromagnetic interference, 167, 179, 353, recording electrode of, 283-284
408 sensitivity of, 291
artifacts due to, 259 single-fiber needle electrode of, 286
Electromagnetic radiation, 109 spectral analysis of, 292
Electromagnetic spectrum, 109 spontaneous activity of, 287
Electromagnetic wave, 521 stimulating electrode of, 283, 284
Electromechanical transducer, 204 surface electrode of, 284
Electromedical device, 167, 306 sweep speed of, 291
Electromyography, 264, 281-293 voluntary activity of, 287
definition of, 264 EMG motor response study
 Index

latency of, 289 Enzyme sensors (see biosensors)

supramaximal stimulation of, 289 EP (see evoked potential)
EMG spontaneous activity, 287 EP studies, signal averaging of, 293
biphasic potential of, 287 Epicardial lead (see also myocardial lead), 341
monophasic potential of, 287 Epilepsy, 263, 265
EMG voluntary activity Epoxy housing, 349
full effort of, 288 Equipment standardization, 370
mild effort of, 287 Equipotential grounding, 196
moderate effort of, 288 Error, 30
EM1 (see electromagnetic interference) absolute, 30
Emissivity, definition of, 464 gross, 30
Endocardial lead, 341,352 random, 30
Endoscope, 529,537 relative, 30
disinfection and sterilization of, 545 systematic, 30
water leakage of, 545 ERV (see expiratory reserve volume!
Endoscopic electrosurgical procedure ESU (see electrosurgical unit)
electrical leakage in, 546 ESU active electrode
patient burn in, 546 ball electrode of, 397
RF leakage current in, 546 flat blade electrode of, 397
secondary site bum in, 546 foot switch of, 397
Endoscopic procedure, 407 loop electrode of, 397
Endoscopic video system, 536-546 multiple use, 397
display monitor of, 538 needle electrode of, 397
endoscope of, 538 single use, 397
image management system of, 538,544 ESU output waveform, 396
image processor of, 538 crest factor of, 395
light source of, 538 ESU pencil, 397
video camera of, 538-539 ESU return electrode
Endoscopy, 537 bum at, 398,407
bleeding in, 545 conductive gel pad, 398
cancer detection in, 544 electrode-skin contact at, 398
perforation in, 545 skin effect in, 398
problems of, 545 tissue damage at, 398
three-dimensional, 544 Ether, 487
wireless camera capsule, 544 Ethernet, 235
Endotracheal tube, 423,481,493 bandwidth of, 235
End-tidal carbon dioxide, 41 1 collision detection of, 235
End-tidal C 0 2 monitor, 480-485, 497-498 Ethylene oxide, 545
airway adaptor of, 483 E-type thermocouple, 80
errors in, 484 Evoked potential, 263-280
functional block diagram of, 484-485 Evoked potential study, 264,282-293,
mainstream, 482-484 282-293
nitrous oxide in, 485 definition of, 263
pressure fluctuation in, 485 Expiration, 411
sidestream, 482, 484 Expiratory reserve volume, 411
water trap of, 484 Explosion hazard, 183
Enflurane, 490 External pacemaker, 352
Enriched oxygen environment, 190,407,531 Extra-low voltage, 198
Enteral feeding, 372 Eye protection, 531
 Biomedical Deuice Technology

universal cord of, 540

water channel of, 540
Fanfold paper, 204 Floating surface electrode, 156
Fast Ethernet, 235 Flow transducer (see also transducer, flow),
FDDI (see Fiber Distributed Data Interface) 415-416
Feedback process, 7 Fluid-in-glass thermometer, 67-68
FEF25-75% (see maximal midexpiratory Force transducer (see also transducer, force),
flow) 54
Fahrenheit temperature scale, 66 Forced expiratory volume, 412
Fetal ECG, 446 Forced vital capacity, 412
Febl heart rate, methods of monitoring. Formaldehyde, 545
445-447 Fourier analysis, 547-550
Fetal heart rate monitoring Fourier series, 547
abdominal ECG method in, 446 Four-wire RTD, 74
direct ECG method in, 446 FOV (see field of view)
phono method in, 446 Fractional oxygen saturation, 471
ultrasound method in, 446 Frequency domain, 36,548
Fetal heart sound, 446 Frequency hopping spread spectrum, 554
Fetal monitoring, 443 Frequency modulation, 554
fetal heart rate in, 444 Frequency response, 46-47
maternal uterine activity in, 444 Frequency spectrum, 36,547,550
Fetal monitors, 443-448 Frontal lobe, 264
Fetoscope, 538 Frontal plane leads, 246-247
FEVl (see forced expiratory volume) FRV (see functional residual volume)
FHR (see fetal heart rate) Fuel cells, 145, 147
FHSS (see frequency hopping spread spec- Fulguration, 393-394
trum) Full bridge, 51
Fiber Distributed Data Interface, 237 Functional building blocks, 9-10
Fiberscope, 540 Functional oxygen saturation, 471, 473
Fibrillation, 356 Functional residual volume, 413
Fick principle, 324-325 Fundamental frequency, 547,548
Fick's Law, 10, 149 FVC (see forced vital capacity)
Field of view, 464, 537
Filters, 39
band pass, 39
band reject, 39 Galvanic cell, 126
high pass, 39 Galvanic oxygen cell, 143
low pass, 39 Gamma curve calibration, 543
Fire hazard, 183 Gas cylinder, 489
Flammable anesthetic agent, 190 dimension of, 555
Flammable gas, 190 stem of, 556
Flexible endoscope, 537-543 yoke of, 556
air channel of, 540 Gas Law, 56
angdations control of, 540 Gauge pressure, 55
control head of, 539,541 Gel-filled electrode, 156
fiber-optic bundle of, 540-541 General anesthesia, 343,418, 487, 496
insertion tube of, 539-542, 545 Generalized epilepsy, 265
instrument channel of, 540 Gezo Jako, 521
light guide connector of, 540-541 GFCI (see ground fault circuit interrupter)
suction channel of, 541 Glucose oxidase, 144
 Index

Glutaraldehyde, 545 Helium dilution method, 413

Grand mal seizures, 265 Hemodialysis, 500-501, 504, 515
Gravity flow infusion, 373 Hemodialysis system
Gravity flow intravenous infusion set, 373 aidfoam detector of, 514
Graybody, 111-112 arterial blood line of, 513
definition of, 464 arterial drip chamber of, 514
near, 464 blood leak detector of, 515
Ground fault circuit interrupter, 197 cleaning and disinfection of, 518-519
Ground reference, 195 contamination of, 519
Grounded power system, 190 dearation chamber of, 514
Guarding shield, 171-172 dearation pump of, 5 14-5 15
dialysate delivery circuit of, 513-515
extracorporeal blood circuit of, 513
fluid flow diagram of, 513
Half bridge, 50 functional block diagram of, 505-506
Half-cell potential, 129 heat exchanger of, 514-515
Half-reaction, 127, 130 heater compartment of, 514
Hall effect. 92 metering chamber of, 514515
Hall effect sensor, 92-94 metering pump of, 514
Hall voltage, 92 patient interface of, 510-51 1
Halothane, 490 pressure sensor of, 514
Hand-switched ESU pencil, 397 roller pump of, 514
Harmonics, 548 ultrafiltration control of, 514
Health care network standards, 238 venous blood line of, 513
Health Level 7,238 water treatment in, 517-518
Heart block, 339 Hemoglobin, 470
1st degree, 339 Hemostatic, 392,395
2nd degree, 339 Heparin, 514
3rd degree, 339 Heparinized saline, 310
Heart rate, 324 High frequency harmonics, 167
Heart stimulation, strength duration curve of, High frequency ventilator, 424
344 HIS (see hospital information system)
Heart Histocompatibility, 22
aortic valve of the, 296,299 HL7 (see Health Level 7)
left atrium of the, 295, 325 Holter ECG (see also ambulatory ECG), 244
left ventricle of the, 295, 299, 325, 330 Homogenous circuit, the law of, 81-82
mitral valve of the, 295 Hospital information system, 230, 544
pulmonary valve of the, 296, 332 Hot air anemometer, 415
right atrium of the, 295,330 Hot film anemometer, 107
right ventricle of the, 295, 325, 330 Hot wire anemometer, 107
tricuspid valve of the, 295, 332 H R (see heart rate)
Heat disinfection, 518 Hub, network, 237
Heat loss Human factor, 14
conduction, 450 Human-machine interface, 14
convection, 450 Humidifier, 430-431
evaporation, 450 Hydrogen electrode, 129
radiation, 450 Hydrogen reference electrode, 135
Heat sensitive paper (see thermal paper) Hydrogen-oxygen fuel cell, 145-146
Heated thermistor, 418 Hypercapnia, 497
Helical coil, 68 Hyperthermic, 459
 Biomedical Device Technology

Hypothalamus, 465 tissue culture method of, 25

Hypothermia, 13, 459 functional, 25
Hypoxia, 444,470, 478, 496 nonfunctional, 25
Hysteresis, 34 Incompressible fluid, 96
Indicator dilution method, 324
indicator recirculation in, 327-328
Indicator electrode, 134
I:E ratio, 428 Indocyanine green
IBP (see invasive blood pressure) nondiffusible indicator of, 327, 333
IC (see inspiratory capacity) Induced current, 167
IC temperature sensor (see integrated circuit Infant incubator, 449-453
temperature sensors) access door of, 450
ICD (see implantable cardiac defibrillator), access port of, 450
340 cuffed ports of, 452
ICHD, 345 disinfection and sterilization of, 452
ICU (see intensive care unit) functional components of, 452
Ideal electrode, characteristics of, 1\51 modes of temperature control, 450
IEC (see International Electrotechnical Com- oxygen controller of, 452
mission) plastic hood of, 450
IEEE 802.11 Standards, 553 proportional heating control of, 451-452
IEEE 802.3 Standards, 235,553 relative humidity of, 451
IEEE 802.5 Standards, 235 water reservoir of, 453
Illuminance, 111 Infant warmer, 450
Impedance matching, 161 Inferior vena cava, 296
Implantable cardiac defibrillator, 340, 357 Infrared spectroscopy, 482
Implantable pacemaker Infrared thermometry, 463-468
acceptable lead impedance of, 348 narrowband, 466
ADL rate of, 348 two color, 466
AOO, 346 wideband, 466
battery monitor of, 350 Infusion controller, 373, 378-379
battery of, 349 Infusion devices, 371-390
chassis of, 349 Infusion pump, 373, 378-390
DDD, 346 air-in-line detection of, 386
DDDR, 346 backpressure of, 389
functional block diagram of, 350 bolus infusion of, 389
hysteresis of, 348 bolus of, 383
lead impedance of, 348 dose-error reduction system of, 386
lower rate of, 348 downsteam occlusion of, 385
output circuit of, 350 flow accuracy of, 389
performance parameters of, 348 flow pattern of, 383-384,388
pulse width of, 348 flow rate of, 380, 385, 388
rate limit of, 348, 350 fluid depletion alarm of, 385
refractory period of, 348 fluid viscosity of, 389
sensing amplifiers of, 350 functional block diagram of, 386-388
sensitivity of, 348 occlusion pressure alarm of, 385
upper sensor rate of, 348 runaway prevention of, 386
VVI, 346 upstream occlusion of, 385
IMV (see intermittent mandatory ventilation) Ink-jet printer, 208
In vivo test, 25 print cartridge, 208
blood contact method of, 25 print head of, 208
 Biomedical Device Technology

Laser action, 522

Laser beam
Labor, 443 reflection of, 532-533
Lambert Beer's Law, 471,473 scattered radiation of, 532
Laminar flow, 96,99 Laser delivery system, 528-530,534
LAN (see local area network), 553 articulation arm of, 528
Laparoscopic procedure, 407 contact laser probe of, 529
Laparoscopy, 537-539 first surface mirror of, 528
Laser, 52 1-535 optical fiber of, 528-529
aiming beam of, 527 total internal reflection in, 529
applications of, 522 Laser plume, 532
carbon dioxide (C02),526-527 Laser power meter, 534
characteristics of medical, 526-527 Laser printer, 203, 206-208
characteristics of, 521 cleaning blade of, 207
Class I, 531 developing cylinder of, 207
Class 11,531 display of, 278
Class IIa, 531 erase lamp of, 207
Class IIIa, 531 laser beam of, 206
Class IIIb, 531, 534 paper speed of, 206
Class IV, 531,534 photosensitive drum of, 206
classification of, 531 primary corona wire of, 207
collimated beam of, 533 scanning mirror of, 206
continuous, 525 toner of, 207
cooling system of, 527 transfer corona wire of, 207
dye, 526,527 Laser procedure
excimer, 526,527 access control in, 532
excitation source of, 522 safety zone in, 532
exposure time of, 524 warning sign in, 532
focal spot of, 525 Laser safety committee, 531,534
functional components of, 526-528 Laser safety hazard
helium neon (HeNe), 526-527 bum, 530,532
holmium:YAG, 527 explosion, 531
KTP/532,527 eye injury, 530-532
lasing medium of, 522,526 fire, 530, 533
ND:YAG, 527 Laser safety officer, 531, 534
optical cavity of, 522 Laser safety program, 534
photo stimulation effect of, 533 administrative control in, 534
power density of, 524-525 engineering control in, 534
pulsed, 525 personal protection in, 534
reflective mirrors of, 522 preventive maintenance in, 534
resonator of, 522 Laser smoke evacuator, 532
ruby, 527 Laser tissue effect, 524-526
semiconductor, 526 ablate, 529
surgical applications of, 524-526 coagulate, 529
surgical effect of, 525 cut, 529
thermal effect of, 533 necrosis zone of, 524
tissue effect of, 524-526 vaporization zone of, 524-525
transport medium of, 527 vaporize, 529
transverse electromagnetic mode of, 526, zone of injury in, 524
534 Lateral strain, 59
 Index

LCD (SMliquid cry~ta!&play) LOW uravefom {SM &O monophaslie

Lead acid cell, 145-146 damped sinusaicfal waveform), 358:
hd m,72-23 Lwr laek c.snnecbrF373
b & w e current, 398 Laid Mvmi, 148
capacitive, 188 L m k c o , 111
mewarement oaF, 199-201 Lurnimws density, 1 l t
resistive, la8 -nous mit.hace, 1 l 1
mwma,m Luminm mew, l I l
Light pip*C55#533;437 l.mninwsflux, 11 I
Light,pal&&on, 212 LumImous, flux density311 1
Ight sausce, 542,543 Luminous intensiiy, 1 11
apeme esf, 542 Luw capacity, 411
a~~C.am&e bri&tnw wntrol uf* 542 Lung wm$lianee, 411
E ~ Q beIRp$dfAL#?
X of, $49 L W (Whaem va.li;thIediffmenjtial b x w
infiaed filter of, 582 farmer)
life span of, 543
mwcufy v q r , 542
mtal hati&, 5%2
output inten&g, 542 Mam~shock,1&5-1818,276,306
qtlartz halagen, 542 Macroshock, chmcterisrics of, 1 8
m a , 5&2 hh%p&.ti6W d hts&xf:am,kY7-1Y9
Li&tniq surge, 1% Umetic nu%* 94
LIM [sw line S~aiatbnmonltor) Manometer, SB
Limb e1=tFadm, %J Manadgravity flew irzftden, 373
Limb lead, 246-247 M w flaw r e , 96
Line bolkxtian monihr, @$-I94 F & . e d ECG, 446
Linear a o c e I e r ~533
, Maternal heart & a d ,446
Linear displmment kanduc~r,$8 Maternal monitering, 444
kXW p&kdk bfW&~ 381-3&
pI%%$l, M m&pha~sq b w , 413
Linear variable differential hmdonner, 98 MDS (sf@m~nophmicdamped sinuscrfdal),
Linearity, 31 358
Liquid ccy.st;rl, 68
Uqu'ud eryscaZ why,Ii2-2UV307,.t5t3
active =ah+, 214
dthsiag elecirade af, 213
ts&-Et of3214
 Biomedical Device Technology

expiratory limb of, 431-432 Monitoring ECG, 243

expired gas of, 430 Monochromatic, 521
flow control of, 431 Monophasic damped sinusoidal waveform,
flow sensor of, 431-433 358,360-362
functional block diagram of, 429-431 Monophasic truncated exponential wave-
high pressure alarm of, 433 form, 358,362
inspiratory limb of, 431 Monopolar operation, 395
inspired gas of, 430 Montage
loss of gas supplies of, 433 electrode selector matrix of, 275-276
loss of power of, 433 montage selector, 274
oxygen supply line of, 431 referential, 275, 277
oxygen/air blender of, 431 Motion transducer (see also transducer, mo-
patient circuit of, 430 tion), 88-89
pneumatic system of, 430-433 Motor unit action potential, 282, 285-286,
power-up self test of, 434 292
pressure cycled, 424 MTE waveform (seemonophasic truncated
pressure sensor of, 431 exponential waveform)
safety backup of, 430 MUAP (seemotor unit action potential)
safety features of, 433-434 Multiparameter monitor, 225
user interface of, 430 Multiple sclerosis, 265
water trap of, 431 Multiprogrammable pacemaker, 340
Y-connection of, 431 Muscle sensing, 353
Mechanics of breathing, 410 Muscle stimulation, 353
Medical air, 429 Myocardial lead, 341, 343
Medical device Myocardium damage, 358
classification of, 5
definition of, 4
front-end of, 52
Medical gas supply system, 555-557 NASPE, 345
wall outlet of, 557 Natural pacemaker, 338-339
Medical instrumentation, 240 definition of, 240
Medical laser, 520-535 NBG, 345,347
Medical telemetry, 551-554 Near graybody, 464
Membrane indicator electrode, 137 Nebulizer, 430-431
Membrane potential, 10, 149 Needle electrode, 157
Membrane resistance, 150 Negative pressure ventilator, 423
Mercury-in-glass thermometer, 67, 458, 462 Nephrophane, 508
Metal indicator electrode, 133-134 Nernst Equation, 132-134
Methemoglobin, 471-472 Nerve conduction velocity, 282
Microshock, 27, 183, 185-188, 306 Nerve potential, 14
characteristics of, 187 Network connection components, 236
sampling rate of, 276 Network interconnection devices, 237
Microsoft Windows, 236 Network interface card, 237
Minimally invasive surgery, 537,543 Network model, 235
Minute volume, 428, 485 client-server, 236
MIS (see minimally invasive surgery) host-terminal, 236
Mixed venous blood, 326 peer-to-peer, 236
Modes of ventilation, 425-427 Network operating system, 236
Modulation, 552-554 Network topology, 233
Monitoring device, 5 bus or tree, 234
 Index

ring, 234 Orthogonal frequency division multiplexing

star, 234 spread spectrum, 554
Networking, 487 Oscillometric NIBP monitor
Neuromuscular transmission time, 29 1 amplifier of, 318
NIBP (see noninvasive blood pressure) analog to digital converter of, 319
NIC (see network interface card) central processing unit of, 319
Nickel metal-hydride cell, 146, 369 display of, 319
Nickel-cadmium cell, 146, 369 functional block diagram of, 319
Nitrogen concentration curve, 414 oscillometric filter of, 318
Nitrogen method, 414 overpressure switch of, 319
Noncontact temperature sensor, 458 pressure sensor of, 318
Non-fade display, 211 printer of, 319
erase bar of, 211 pump and solenoid valve of, 318
waveform parade of, 211 watchdog timer of, 319
Noninductive resistor, 405 Osmosis, 501, 517
Noninvasive blood pressure measurement Osmotic pressure, 502
auscultatory method of, 313-316 Oxygen analyzer, 326,478-479,496-498
oscillometric method of, 313, 317-320 Oxygen cost of breathing, 412
Noninvasive blood pressure monitor, Oxygen permeable membrane, 142
312-322 Oxygen saturation, 14
Noninvasive external pacemaker, 352 Oxygen sensor, 430
Noninvasive method, 45 Oxygenated blood, 296,325
Nonisolated output, 367 Oxyhemoglobin, 470-473
Nonlinearity, 32
Nonthrombogenic surface, 23
NOS (see network operating system)
Notch filter, 39, 257, 272, 278 Pacemaker analyzer, 343
Novel1 Netware, 236 Pacemaker programmer, 340
Pacemaker, 338
asynchronous mode, 340
demand mode, 340
Obstructive lung disease, 412 external noninvasive, 339
Occipital lobe, 264 external invasive, 339
Ocular exposure, 531 implantable, 339
OFDMSS (see orthogonal frequency division lead system of, 339
multiplexing spread spectrum) modes of operation of, 340
Operating range, 69 pacing rate of, 340
Operational amplifier,l59 problems with, 353
bandwidth of, 159 pulse amplitude of, 340
ideal characteristics of 159 pulse duration of, 340
input impedance of, 159 pulse generator of, 339, 341, 343
open loop gain of, 159 rate-modulated mode, 340
output impedance of, 159 sensitivity of, 340
Opitcal densitometer, 333 transcutaneous, 339
Optical intensity ratio, 474, 476 Pacing current, 341
Optical isolator, 195 Pacing threshold, 344
Optical path length, 471, 473, 477-478 Packet transmission, 553
Optical transducer (see also transducer, opti- Paper chart assembly, 204
cal), 43 Paper chart recorder, 203-210
Orifice plate, 101 continuous paper feed, 203
 Biomedical Device Technology

ink-jet, 203 Photoconductive sensor, 114

paper supply mechanism, 204 Photodetector, 110
single page feed, 203 Photodiode, 115-1 18
thermal dot array, 203 photoconductive mode of, 116
thermal stylus of, 203 photovoltaic mode of, 116
Paradoxical sleep, alpha wave of, 272-273 Photodynamic drug, 533
Parenteral nutrition, 372 Photoelectric effect, 115
Particle drift, 10 Photoelectric tube, 115
Passive breathing, 410 Photoemissive sensor, 115
Passive electrode, 392, 397 Photometry, 109
Patient breathing circuit, 478-479 Photon detector, 110
circle system of, 492 Photon, 521
expiratory limb of, 493 Photoresistive sensor, 113
inspiratory limb of, 431, 478, 493-494 Phototherapy light source, life span of, 455
T-piece design of, 492, 494-495 Phototherapy light, 450,453-456
Patient interface, 25 dehydration in, 454
Patient leakage current, 189 far-infrared radiation in, 454
Patient monitoring network, 227, 233 fluorescent tube of, 454
desired characteristics of, 233 height adjustable stand of, 454
Patient-initiated mandatory breath, 425 hyperthermia in, 454
Paul M. Zoll, 338 observation timer of, 454
PAWP (see pulmonary arterial wedge pres- spectral output of, 454, 455
sure) ultraviolet radiation in, 454
p C 0 2 electrode, 141, 142 UV filter of, 454
PCW (see pulmonary capillary pressure) Phototransistor, 118
PEEP (see positive end-expiratory pressure) Physiological and tissue effects of electrical
Perception limitation, 15 current, 184
Perception of pain, 184 Physiological monitor, 224
Percutaneous venous cannula, 511 alarm function of, 225
Periodical signal, 36 analyze function of, 225
Peripheral temperature, 459 condition function of, 225
Peristaltic pumping mechanism, 373, display function of, 225
379-382 monitored parameters of, 226
protruding finger of, 380 record function of, 225
rotating cam shaft of, 381 sense function of, 225
Peritoneal cavity, 515-517,537,544 Physiological monitoring, 481, 487, 490
Peritoneal dialysis, 500, 504, 515-517 Physiological parameters, 14, 224, 487, 551
continuous ambulatory, 517 Physiological signals, 13-14, 324
continuous cycler-assisted, 516 characteristics of, 14
Peritoneal membrane, 504, 515 Piezoelectric constant, 62
Peritonitis, 517 Piezoelectric pressure transducer, 62
Permeability, 89 Piggyback infusion, 375
Permittivity, 9 1 PIM (see patient-initiated mandatory breath)
Petit ma1 seizures, 265 Pin-indexed safety system, 497-498, 556
Petrolate-in-glass thermometer, 67 Piped-in wall outlets, 555
pH electrode, 140-141 Piston cylinder pumping mechanism, 373,
Phase distortion, 46, 48 379-380
Philip Drinker, 423 cam of, 379
Photocathode, 115 input port of, 380
Photocoagulator, 525 output port of, 380
 Index

stepper motor of, 379 Primary reservoir, 555

stroke volume of, 380 Programmable pacemaker, 340
valve of, 380 Protective eyewear, 532
Pitot tube, 101 Pseudoperiodical, 36
pK electrode, 139-140 PSG (see polysomnography), 263, 264
Placental blood flow, 444 Pulmonary arterial wedge pressure, 333
Planck's constant, 109 Pulmonary artery, 296,325,330,459
Planck's law, 463 Pulmonary capillary pressure (see also pul-
Plasma display, 215 monary arterial wedge pressure),
brightness of, 217 333
contrast ratio of, 2 17 Pulmonary function lab, 412
refreshing rate of, 217 Pulmonary vein, 296
resolution of, 217 Pulse oximeter, 460-479,496-498
viewing angle, 2 17 dark signal of, 475
Plexiglas, 454 functional block diagram of, 475
P-N junction, 116 light-emitting diode of, 475
Pneumothorax, 410 photodetector of, 475
p 0 2 electrode, 142-143 plethysmograph of, 476
Poiseuille's law, 99 reflecting probe of, 475
Poisson's ratio, 59 sensor probe of, 475
Polarization, 240-241 signal to noise ratio, 477
Polarized cell, 149 timing control circuit of, 475
Polarographic cell, 142 transmitting probe of, 475
Polysomnography, 263-264 , Pulsed laser, 525
definition of, 264 Pumping mechanism, 373
Polytetrafluroethylene, 510 Purkinje fiber, 338
Pons and medulla, 410 Purkinje system, 240
Poor perfusion, 477 P-vectorcardiogram, 253
Population inversion, 521 Pyroelectric sensor, 113-114,466
Portable ventilator, 424 ferroelectric material of, 466
Positive end-expiratory pressure, 426 Pyrometry, 463
Positive pressure ventilator, 423
flow cycled, 424
time cycled, 424
volume cycled, 424 QRSvectorcardiogram
waveform of, 424 esophageal lead of, 252
Potentiometry, 142 Quantum efficiency, 123
Power line filter, 179 Quantum event, 109-1 10
Power line interference, 169
Power noise, 60 Hz, 166
Precision, 30
Precordial lead (see also chest lead), 247 Radiance, 111
Premature ventricular contraction, 243 Radiant density, 110-1 11
Pressure and force transducer, 53-64 Radiant energy, 110- 111
Pressure support, 426 Radiant flux, 110-1 11
Pressure transducer, 44,56-57, Radiant flux density, 110- 111
61, 63,300-305,311, 316, Radiant intensity, 111
318,333,447-448 Radio frequency, 392
Prevost and Batelli, 356 Radiometry, 109
Primary cell, 145 Radiopaque, 349
 Biomedical Deuice Technology

Radiant emittance, 110-11 1 Return electrode, 341, 392-393, 407

Rate-modulated pacemaker (see rate-respon- Reusable trailsducer, blood pressure, 305
sive pacemaker), 346 Reverse bias current, 116
Rate-responsive pacemaker, 338 Reverse osmosis, 518
Redox reaction, 126-127 Reynolds number, 99
Reference electrode, 134-138 RF (see radio frequency)
definition of, 135 RF current density, tissue effect of, 393
Refractory period, 150, 356, 368 Right atrium, 342
REM (see return electrode monitoring) Right ventricle, 342
Remote monitoring station, 226 Right-leg-driven circuit, 173-177
Renal deficiency, 504 electrode placement of, 247
Renal dialysis, 500, 517 Rigid endoscope, 538-539,542
concentration gradient in, 501-502,505 eye piece of, 538-539
double-needle technique in, 51 1 instrument channel of, 538-539
mass diffusion rate of, 501 light cable of, 538
mechanism of, 504-505 light source of, 538
rate of diffusion in, 501 optical fiber of, 538
single-need technique in, 5 11 rod lens of, 538
Repeater network, 237 Risk classification, device, 7
Repolarization, 241 Risk current, 188, 199, 306
membrane, 11 RLD circuit (see right-leg-driven circuit)
Reproducibility, 31 Roller clamp (see also regulating clamp of IV
REQM (see return electrode quality monitor- set), 377
ing) Roller pump (see rotary peristaltic infusion
Reserve supply, 555 P ~ P )
Residual volume, 411 Rotameter, 103
Resistance temperature device, 69-72, 86 Rotary peristaltic infusion pump, 381
characteristics of, 72 Router, 237
Resistivity, 88, 392 Routing protocols, 238
Resolution, 30 Routing table, 238
Respiration monitor, 409-421 RTD (see resistance temperature device)
amplitude modulated signal in, 419 RV (see residual volume)
functional block diagram of, 420
lead selector of, 421
synchronous demodulator of, 421
Respiration monitoring SA node (see sinoatrial node)
heated thermistor method of, 418 Saline, 328
impedance pneumographic method of, Salt bridge, 128
418-421 Saturation, 34
muscle and nerve stimulation in, 419 Scalp electrode, 270-272
Respiration rate, 14, 41 1 electrode placement of, 270-272
Respiratory arrest, 487 Scavenging system
Response time, 33, 69 passive exhaust, 495
Resting potential, 149 vacuum, 495
membrane, 11 wall suction of, 495
Restrictive lung disease, 413 Scotoma, 531
Retrolental fibroplasia, 452 Screw pump (see syringe infusion pump)
Return electrode monitoring, 399-400 Sealed lead acid cell, 369
Return electrode quality monitoring, Secondary burn, 366
399-400 Secondary cell, 145
 Index

Secondary reservoir, 555 Sodium-potassium pump, 10, 149

Seebeck coefficient, 80 Solar cell, 116-1 17
Seebeck effect, 80 Solenoid, 90
Seebeck voltage, 80 Spark-gap generator, 392, 401
Selective membrane, 507 Specifications, 29
active permeability of, 507 Spectral irradiance, 453
passive permeability of, 507 Sphygmomanometer, 313-314, 316
Selective radiator, 111, 112 cuff of, 314-315
Self-heating error, 75 hand pump of, 314
Semipermeable membrane mercury manometer of, 314
permeability of, 501 pressure measurement device of, 314
Sensing threshold, 344 rubber bladder of, 313
Sensitivity, 31 valve of, 314
Sensitivity drift, 31 Spiral electrode, 343, 446
Sensor, 43 Spirometer, 415-417
Sensor-indicated interval, 347 bellow of, 415
Sensory limitation, 15 flow transducer of, 415
Sensory nerve action potential, 290 flow-sensing, 415
Servoflurane, 490 functional block diagram of, 415
Shell temperature (see peripheral tempera- volume transducer of, 415
ture), 459 volume-sensing, 415
Shielded lead, 170 water-sealed inverted bell, 415
Shock prevention methods, summary of, 199 Spontaneous breath, 425
Sigh, 427 Spore-forming bacteria, 518
Signal conditioning, 48 Spread spectrum technology, 553
Signal extraction, 477 Stability, 69
Signal isolation, 26, 195-198 Standard atmospheric pressure, 54
demodulator, 196 Standard Electrode potential, 130-131
modulator, 196 Standard oxidation potential, 130-131
Silver/silver chloride electrode, 154-155 Standard reduction potential, 130
electrical equivalent circuit of, 154-155 Standard reference electrode, 129
Silver/silver chloride reference electrode, Statistical control, 33
135-136 Steady state characteristics, 33
SIMV (see synchronous Intermittent manda- Stephen Hales, 295
tory ventilation) Stethoscope, 3 14
Single forced expiration, 412 Stimulated emission, 521
Single page feed recorder, 205 Storage oscilloscope, 211
Single-cell membrane potential, 10, 149 Strain gauge, 58
Single-use disposable device, 377 bonded, 61
Sinoatrial node, 240, 338 diaphragm, 57-62
Sinus bradycardia, 339 gauge factor of, 60
Sinus rhythm, 338, 356 metal wire, 61
Sinus tachycardia, 339 piezoresistive, 61
Sinusoidal signal, 36 unbonded, 61
Skin electrode, 154, 446 Stray capacitance, 188
Skin preparation, 170 Streamline flow, 96
Skin temperature, 450, 458 Strength duration curve, 344
Sleep disorder, somatosensory, Stress test, 243
264 Stress, 24
Sodium hypochlorite, 518 Stroke volume, 324
 Biomedical Device Technology

S-type thermocouple, 81 laser)

Subclavian vein, 343 Temperature, body, 14
Subsystems, 7 Temperature coefficient, 72, 76-78
Superior vena cava, 296,338 Temperature monitor, 458
Surface electrode, 154 excitation circuit of, 462
Surge protector, 179 functional block diagram of, 462
Surgical instrument, 521, 537 Temperature sensitivity, 72
SV (see stroke volume) Temperature sensor (see also temperature
Swan-Ganz catheter, 324,330-332 transducer), 458
balloon of, 332 Temperature transducer (see also transducer,
distal lumen of, 332 temperature), 76, 79, 459-460
injectate orifice of, 331 electrical, 69-86
injectate port of, 332 Temporal lobe, 264-265
thermistor of, 331 Temporary cardiac pacing, 351
Switches, network, 237 T I T (see thin film transistor)
Switching power supply, 167 Therapeutic device, 6,26,338
Switching transient, 167, 179 Thermal detector, 110
Synchronous cardioversion (see also cardio- Thermal dilution curve, 329-330,332-333
version), 356 Thermal dilution method, 328-330
Synchronous Intermittent mandatory ventila- bolus of injectate of, 335
tion, 426 catheter dead space of, 334
Syringe infusion pump, 373,379,382-384 frequency of injection of, 335
System boundary, 8 injectate recirculation of, 335
System input, 7 injectate temperature of, 335
System output, 7 injectate volume of, 335
System, 7 injectate warming of, 334
closed, 7 intravenous administration of, 335
open, 7 rate of injection of, 335
Systems Approach, 7-8, 17, 19 thermistor position of, 335
Systolic blood pressure, 296, timing of injection of, 334
298-299,307,314-317,320,322 Thermal dot array recorder, 204-205
print head of, 204
vertical resolution of, 205
Thermal event, 109
T-1 lines, 237 Thermal flowmeter, 106-107
TCP/IP, 235 Thermal paper, 203
Teflon encapsulated, 67 Thermionic electrons, 114
Telemetry, 230, 340, 551 Thermistor, 69, 76-79, 86, 459, 460
bandwidth requirement of, 552 Thermistor linearization, 78
channel overcrowding of, 552 Thermocouple, 69, 79-86, 113
data rate of, 552 cold junction of, 80
EM interference and immunity of, 552 exposed, 85
power requirement of, 552 grounded, 85
primary user in, 552 hot junction of, 80
receiver of, 230-231 thermoelectric sensitivity of, 81
reliability of, 552 ungrounded, 85
secondary user in, 552 Thermometry, 458
transmission range of, 552 Thermopile, 113-1 14,466
transmitter of, 230-231 thermocouple in, 466
TEM (see transverse electromagnetic mode of Thin film transistor, 214
 Index

Thomas Seebeck, 79 detector of, 465

Three-wired RTD, 73 errors in, 467-468
Threshold drift, 353 filter of, 465
Threshold of perception, 184, 187 FOV of, 465
Tidal volume, 14,411,426 functional block diagram of, 465
Time constant, 69 multiple scanning of, 466
Time domain, 36, 548 offset of, 467
Time-varying signal, 36 optical lens of, 465
Tissue effect, 183, 392 optical light pipe of, 465
Tissue-electrode interface, 152, 154-155 shutter mechanism of, 466
Titanium housing, 349 slgnal processor of, 467
TLC (see total lung capacity) thermistor of, 466
TOCO transducer, 447
Token Ring
bandwidth of, 235
characteristics of, 235 U.S. Food and Drug Administration, 338
multistation access unit of, 235 UA (see uterine activity)
Total lung capacity, 41 1 UHF, 552-553
Trabeculae, 344 Ultrafiltration coefficient, 507
Tracheostomy tube, 423 Ultrafiltration, 501-502, 505, 509, 511, 515
Transcutaneous pacemaker, 352 Ultrasound blood flow detector, 435-442
Transducer, 43-157 Ultrasound Doppler flowmeter, 104
active, 44 Ultrasound flowmeter, 104-105
definition of, 43 Ultrasound receiver, 438
direct mode of, 44 Ultrasound transmitter, 438
electrochemical, 125- 147 Ultrasound vortex flowmeter, 105
excitation, 49 Ultrasound
flow, 95-107 frequency of, 436
force, 53-64 heating effect of, 436
indirect mode of, 44 propagation speed (or velocity) of, 436
motion, 87-9 4 wavelength of, 436
optical, 108-124 Ultrasonic cleaner, 545
passive, 44 Unipolar lead, 341-342, 353
pressure, 53-64 UNIX, 236
temperature, 65-86 User fatigue, 17
Transduction element, 45 User interface, 26
Transfer function, 38 Uterine activity, 444-445
Transient characteristics, 33 amplitude of, 445
Transit time flowmeter, 437 characteristics of, 445
Transmission link, twisted copper wire, 236 duration of, 445
Transthoracic pacemaker (see transcutaneous frequency of, 445
pacemaker) methods of monitoring, 447-448
Trocar, 537-538 resting tone of, 445
T-type thermocouple, 81 rhythm of, 445
Turbine flowmeter, 104 Uterine activity monitoring
Turbulent flow, 96, 99 external pressure transducer
TV (see tidal volume) method in, 447
T-vectorcardiogram, 253 intrauterine pressure method in, 448
Tympanic membrane, 459, 465 Uterine contraction, 443
Tympanic thermometer, 458, 463, 465 Ultrasound transit time flowmeter, 105
 Biomedical Device Technology

Volume of dead space air, 414

Volume-to-be-infuse,385-387
Valinomycin, 139 Volumetric infusion pump, 373
Vaporizer, 489-49 1 VTBI (see volume-to-be-infused of infusion
agent chamber of, 491 pump)
concentration control of, 491
control valve of, 491
electronically controlled, 490-49 1
keyed filling spout of, 497-498 Wall piped gas outlets, 488
mixing chamber of, 49 0-49 1 Walton Lillehei, 338
reservoir of, 491 WAN (see wide area network)
temperature-sensing flow control of, 491 Waste anesthetic gas, 493, 495
vaporizing chamber of, 490-491 Water softener, 518
variable-bypass, 490 Water treatment, 512
Vascular access, 510-51 1,513-514 Waterborne bacteria, 518
site survival rate of, 511 Weaning, 426
Vascular restriction, 435 Wheatstone bridge, 49, 462
VC (see vital capacity) excitation voltage of, 49
Vectorcardiogram, 247,256 White balance, 543
Velocity, 88 Wide area network, 233, 259
Venous access, 373 backbone of, 237
Venous blood, 325, 410 Wien's displacement law, 463
Ventilation parameters, 427-429, 487 Wilson Greatbatch, 338
Ventilator, 423 Wilson network, 249,253
types of, 423-424 Wireless link, infrared, 236
Ventilator-initiated mandatory breath, 425 WLAN, 553
Ventricle, 240 WMTS, 552-553
Ventricular fibrillation, 183-184, 243, 340, Work of breathing, 412
356-357,366
Venturi tube, 100
VHF, 552
Video processor, 543 Xerox, 235
Videoscope, 540 X-ray fluoroscopy, 330
CCD camera of, 543 Yellow Spring Instrument, 76, 459
mosiac color filter of, 543 YSI (see Yellow Spring Instrument)
rotating color wheel of, 543 YSI 400 probe, 459
VIM (see ventilator-initiated mandatory characteristics of, 459-460
breath), 426 YSI 400 series thermistor, 76
Viscosity, 97-99 YSI 700 probe, 460
coefficient of, 98-99 characteristics of, 461
Visking, 508 YSI 700 series thermistor, 78
Visoelasticity, 24
Visual display monitor, 208-217 z
Vital capacity, 4 11
Vital sign, 487, 498, 551 Zero drift, 31
Voltage limiting device, 180-181 Zero offset, 31
Voltaic cell, 126 Zinc-air cell, 146
Volume flow rate, 96 Zinc-carbon cell, 146
For many years, the tools available to physicianswere limited to
a few simple hand-pieces such as stethoscupes, thermometers
and syringes; medical professionals primarily relied on their
senses and skills to perform diagnosis and disease mitigation.
Today, diagnosis of medical problems is heavily dependent on
the analysis of information made available by sophisticated
medical machineries such as electrocardiographs, ultrasound
scanners and laboratory analyzers. Patient treatments often in-
volve specialized equipment such as cardiac pacemakers and
eiectrosurgical units. Such biomedii instrumentations play a
critiial and indispensable role in modern day medicine. In order
to design, build, maintain and effectively deploy medical de-
vices, one needs to understand not only their design and con-
struction but also how they interact with the human body. This
book provides a comprehensive approach to studying the princi-
ples and design of biomedical devices as well as their applica-
tions in medicine. tt is written for engineers and technologists
who are interested in understanding the principles, design and
applications of medical device technology. The book is also in-
tended to be used as a textbook or reference for biomedical de-
vice technology courses in universities and colleges.