Medical Imaging Modalities
An Introduction
Stephan Scheidegger,
2016
Medical Imaging Modalities ‐ An Introduction
Contents
GRUNDLAGEN RADIOLOGIE
IK
EC HN
Imaging – purpose?
STRAHLENBIOLOGIE
STRAHLENPHYSIK
GEN T
Xrays
ROENT
XRF / XRA
CT
NUK: PET / SPECT
MRI
Ultrasound
Image viewing
Medical Imaging Modalities ‐ An Introduction
Aim
GRUNDLAGEN RADIOLOGIE
IK
You know the basic principle
EC HN
STRAHLENBIOLOGIE
STRAHLENPHYSIK
GEN T
and some clinical
ROENT applications of XR, XRF, XRA,
CT, SPECT, PET, MRI, SONO /
US
Imaging ‐ Purpose?
Different questions – different modalities:
Diagnostic: Imaging of anatomy and physiology!
Image guided therapy: Interventional radiology and
radiotherapy
Theranostics: Tracer with isotpes for functional imaging and
therapy
Xrays – toward an insight into pationts anatomy
Steps to modern diagnostics:
1895: Discovery of Xrays by
C.W. Röntgen
End of 19th Century: first
medical applications
(diagnostics)
Early 20th Century: first
therapeutic applications
1969 / 72 first CTs
(Hounsfield, first conercial
scanner by EMI)
Xrays – toward an insight into pationts anatomy
Anode Principles of Xray generation:
High voltage (U) between
+ anode and cathode (40‐140
kV for diognostics; 20‐300
kV for therapy)
U Accelaration of electrons
between cathode and anode
Electron transport toward
- anode (tube current, mA)
Collision of electrons with
annode (Bremsstrahlung,
Kathode characteristic lines)
The Xray Source
Tube Voltage
Tube
Heating
Curcuit
Hip Imaging …
Koenig et al. 2001, Am J Roentgenol.
after coronar
Angiography
after CT Angiography
Wagner, 2007:Biomed Imaging Interv J
100 mSv 4D-CT
Nieren-Angio
50 mSv
131I-Therapie Cerebr.-Angio
20 mSv
Nephrostomie
10 mSv CT
Abdomen-
5 mSv Thorax
CT Schädel
Skelet Szinti natürliche
2 mSv jährliche
Exposition
1 mSv
LWS a.p.
0.5 mSv Mammographie cc & mlo
0.2 mSv
DVT
Schädel a.p.
0.1 mSv
Zahnstatus
0.05 mSv Thorax p.a.
0.02 mSv Einzelzahn
Conventional Xray – Projection Radiography
What can be seen?
(densities):
Air
Fat
Muscle
Bone
Metal
High contrast
produced by
Photo effect
Conventional Xray – Projection Radiography
Indications):
Bone fractures
Reposition (e.g.
shoulder)
Chest
Detectors
Task: Convert Xrays into a signal:
Photographic system: Film
with fluorescent screen
CR‐system: Imaging plate
DR‐system: Conversion of Xray
quanta into a signal (electrical
charge)
CdTe‐Detektor
Detectors
Information width:
Optical density Small for photographic system
Gray scale (film with fluorescent screen)
Large for CR – and DR‐system:
typical resolution 12 bit,
special monitors for displaying
medical Xray images 10 bit
Radiation dose
XRF & XRA: Fluoroscopy and Angiography
Indications / Applications:
Imaging movements (Hip
and Knee)
Imaging vessels by using
contrast media (cerebral or
coronar angiography)
Interventional radiology
(stents, treatment of
aneurysm)
Medical disciplines:
Neuroradiology, Cardiology,
Urology, Gastro‐Enterology,
Orthopedics, Surgery, etc.)
XRF:
Flouroscopy
XRA:
Angiography (+ Contrast media)
DSA:
Digital Subtractions‐Angiography
(Contrast‐enhanced image – native image)
Cardangio setting
(Dose rate measured up to 2 mSv/h)
Something special?
CT: Computer Tomography
Applications / Indications:
2D‐ (slice) or 3D‐ view
(renederd contours)
available
DVT: Dental Volume
Tomography and CBCT
Soft tissue contrast clearly
better than with
conventional (projection‐)
radiography
CT angiography
Fast reconstruction
algorithm allowing 4D‐CT
(tracking)
RT planning
CT ‐ based Radiation Therapy Planning
EMI‐Scanner
1972
IAEA HUMAN HEALTH REPORTS No. 5: STATUS OF COMPUTED TOMOGRAPHY DOSIMETRY FOR WIDE CONE BEAM SCANNERS
Dental Volume Tomography DVT
CT: 2D or 3D?
CT: How To Get An Image (Slice)
Basic Principle:
Projection: Absorption of
radiation can be detected
as relief
Reconstruct 2‐dim. Image
from n 1‐dim projections
n ∞ for ideal image (n >
36 for real image)
Image matrix
80x80 Pixles
(1974)
Image matrix
512x512 Pixles
(2000)
imbie.meb.uni‐bonn.de/epileptologie/staff/lehnertz/CT1.pdf
CT: How To Get An Image (Slice)
Basic Principle:
Real image is a pixelized
image
Every pixel represents a
grey scale corresponding to
the tissue density
(resolution 12 bit)
Task: find the density of a
pixel based on n
projections
CT: How To Get An Image (Slice)
Basic Principle:
Task: find the density of a
pixel based on n
projections
The absorption
contribution from a specific
direction to every pixel is
defined by the sum of
contributions of every path
element
CT: How To Get An Image (Slice)
Basic Principle:
ds ( x , y )
I I 0e s
The absorption
contribution from a specific
I0
P( ) ln direction to every pixel is
defined by the sum of
I contributions of every path
element
ds ( x, y ) The signal strength P in a
s projection is given by the
ik
the detected intensity I
For absorption, the Beer‐
i Lambert law is assumed
CT: How To Get An Image (Slice)
Algebraic approach:
All projections are defining
a system of (linear)
equations
The projection values P()
are known
Rearrange equations to
find the contributions x
Reconstruction Algoritms
Different approaches:
Arithmetic (slow)
FFT: A projection in the real space
corresponds in the k‐space
(frequency domain) to a slice; FFT of
projection, add all k‐space slices,
reverse FFT
Filtered back projection:
Convolution of real space
projections with filter function (e.g.
Shepp Logan Kerner)
Iterative reconstruction: estimation
of the real pixel value by a statistical
/ physical model
Scanning‐Techniques
Scanning‐Techniques
Pencil-Beam-Geometry
• Parallel beams needed for
reconstruction
Scanning‐Techniques
Fan-Beam-Geometry
• Divergence in slice
Scanning‐Techniques
Fan-Beam-Geometry
• Divergence in slice
• Parallel beams (rays) in
different projections
Scanning‐Techniques
Fan-Beam-Geometry
• Divergence in slice
• Parallel beams (rays) in
different projections
• Reassembly of parallel
beams
Scanning‐Techniques: Axial / Sequential vs. Spiral Scans
CT: Image Content and Window
Hounsfield‐Units x Wasser
HU 1000
Wasser
200
Knochen
Leber
100
Herz
Darm
Water 0
-100
Fett
Lunge
-200
CT: Image Content and Window
Window
+148
+1000 + 120
+ 800 + 100
+ 600 + 80 Displayed
+ 400 + 60
+ 200 Center
grey scale
+ 40
0 + 20 8‐10 bit
- 200 - 0
- 400 - 20
- 600 - 40
- 800 - 60
- 1000 - 80
- 100
HU - 108
soft tissue and lung window
W 400 HU W 750 HU
C +45 HU C -720 HU
PET & SPECT: Tomography in Nuclear Medicine
Applications / Indications:
Metabolic information
(tracer priciple)
19F‐FDG for PET brain
imaging or cancer
metastasis search
99mTcO4 for bone micro
fractions or metastasis
search
Cardiology
Theranostics
PET & SPECT: Tomography in Nuclear Medicine
Priciple opf tracer and imaging:
Tracer (specific molecule defines
biokinetics)
Tracer accumulates in certain structures
(e.g. metastasis or activ brain region)
Isotope defines radiation energy and
characteristics: For PET, a positron emitter
has to be used
PET & SPECT: Tomography in Nuclear Medicine
Typical isotopes
SPECT (Single Photon Emission Computer Tomography)
• 99mTc: phys. HWZ 6.02 h
• 123I: phys. HWZ 13.2 h, biol. HWZ 80 d
• 125I: phys. HWZ 60.14 d, biol. HWZ 80 d
• 131I: phys. HWZ 8.04 d, biol. HWZ 80 d
PET (Positron Emission Computer Tomography)
• 11C: HWZ 20.3 min
• 13N: HWZ 10 min
• 18F: HWZ 109.7 min
Isotopes in use (%, FOPH statistics)
45
40
35
30
25
20
15
10
0
PET & SPECT: Tomography in Nuclear Medicine
How to get an Image?:
Radiation detection with a szintillation
detector
Gamma (Anger‐) cam: Collimation of
radiation (scatter = bluring; no or less
collimation needed for PET)
For tomography: measurements at different
angles / posistions Projections!
SPECT: Single Photon Emission Tomography
How to get an Image?:
For tomography: measurements at different
angles / posistions Projections!
SPECT‐CT
PET: Positron Emission Tomography
How to get an Image?:
+ ‐decay: Emission of
a positron (e+)
Anihilation: two
photons are emitted
The angle between the
photons is (approx.)
180°
PET: Positron Emission Tomography
How to get an Image?:
The angle between the photons is (approx.)
180°
Coincidence: Two photons detected within a
certain time window have the same origin
Position of decay is defined!
The two photons define a projection
For tomography: measurements at different
angles / posistions Projections!
PET: Positron Emission Tomography
CT PET Detector
Image Fusion:
PET and MRI
MRI: Magnetic Resonance Imaging
Applications / Indications:
Soft tissue (not bones due
to lack of signal)
Anatomic and functional /
physiological information
fMRI, blood flow
Spectroscopy
Cerebral diagnostics
Cardiology
…
MRI: Magnetic Resonance Imaging
Principle:
Need of a signal from the
body, which can be located
Water is widely distributed
in the body, concentration
depends on tissue:
Protons have a nuclear
spin, resulting in a
magnetic dipole
Figures by Berberat J (2016), KSA
MRI: Magnetic Resonance Imaging
Principle:
Protons have a nuclear
spin, resulting in a
magnetic dipole
In a thermodynamic
equilibrium without
external magnetic field,
there is no net
magnetisation
Figures by Berberat J (2016), KSA
MRI: Magnetic Resonance Imaging
Principle:
N
with external magnetic
field B0, protons start to
precess
The rate of precession is
the Larmor‐frequency (with
the gyro‐magnetic ratio
for 1H):
B0
B0
S
42.57 MHz / T
Figures by Berberat J (2016), KSA
MRI: Magnetic Resonance Imaging
Principle:
with external magnetic
B0 field B0, protons start to
precess
In a thermodynamic
equilibrium, a net
magnetisation results
(temperature dependent,
given by Boltzmann
equation)
Figures: byBerberat J (2016), KSA
Inside the magnetic field, net magnetization will occur
z
z
B0 B0
y
y
M
x
x
Berberat J (2016), KSA net magnetization
MRI: Magnetic Resonance Imaging
Principle:
The nuclear spins can be
influenced by external
magnetic and electro‐
magnetic fields
Applying a RF‐puls can turn
the spins and subsequent
the orientation of the net
magnetisation
Bloch equations
Magnetic resonance
Energy is applied as radiofrequency (RF) energy
– usually in so-called 90º and 180º pulses (SE)
RF-energy
applied
re-emitting the
M M absorbed energy
(FID)
protons begin to realign themselves to the direction of main magnetic
field Berberat J (2016), KSA
Relaxation
• Decay ofr signal: T2 relaxation, spin-spin or transverse relaxation (xy-
plane)
– decrease of transverse coherence of protons
– energy is exchanged between spins
– sensitive to water mobility
• combination of magnetic field inhomogeneities and spin spin transverse
relaxation, with the result of rapid loss in transverse magnetization and
MRI signal: Free Induction Decay (FID)
Mxy
T2* = total relaxation time
T2= spin-spin relaxation
T2`= component of T2
T2
Relaxation time induced
T2 *
Berberat J (2016), KSA by field inhomogeneities
Berberat J (2016), KSA
T1 relaxation T2 relaxation
increase of longitudinal decrease of transverse
magnetization magnetization
Mz Mxy
T1
T2
t
t
or:
MRI: Magnetic Resonance Imaging
How to get an image?:
Larmor frequency is
dependent on external
magnetic field
A gradient field leads to a
space dependent Larmor
frequency
Emmitted RF‐signal can be
located by the frequency
Figures by Berberat J (2016), KSA
Input: Puls Sequ
Output: (FID) Signal
Gradients
• Three physical gradients: x, y and z gradients
– embedded inside magnet
– used to modify static magnetic field
• Gradients used in imaging
– Slice selection gradient (GSS)
– Read-out or frequency encoding gradient (GRO)
– Phase encoding gradient (GPE)
Berberat J (2016), KSA
From measured signals (frequency and phase map),
regular image can be calculated using Fourier
Transform
freq
phase
Each pixel in the image is related to the amount of
spins and the magnetic environment at the
corresponding location in the sample
In a MR image, different measures can be converted to grey scale and
displayed: Weighting in different images
PDw T1w T2w
TR/TE=5500/14ms TR/TE=500/10ms TR/TE=5500/101ms
SONO/US: Ultra Sound
Applications / Indications:
Soft tissue (not bones due
to high impedance)
Endosonography
Echocardiography
Doppler sonography
Vascular system by using
contrast media
Advantage: non‐ionizing
radiation!
… and gives 2‐dim. slices
through patients anatomy
Endosonography
SONO/US: Ultra Sound
Priciple of image generation:
Mechanical waves emitted
into patients body
Frequency 1‐20 MHz
Reflected waves are
detected
SONO/US: Ultra Sound
Priciple of image generation:
E
c Wave velocity depends on
material (elastical proprieties
and density)
Fatty tissue: c = 1475 m/s
Blood: c = 1570 m/s
Brain: c =1560 m/s
1
c Liver: c = 1570 m/s
Muscles: c = 1580 m/s
Bone (Compacta): c = 3600
m/s
SONO/US: Ultra Sound
Z c Priciple of image generation:
Accoustic impedance Z defines
reflection
I ref Difference of c large R
R large
I inc R large strong reflection
Z 2 Z1
Z 2 Z1
Material Density C ImpedanceN
g/m³ m/s s/m³
Air 1,29 331 429
Water 998 1483 1.48×106
Fatty tissue 920 1410–1479 1.33×106
Bone 1380-1810 2700-4100 4.3-6.6×106
Kramme (Ed.), Medizintechnik, Kapitel 27
SONO/US: Ultra Sound
Transducer
= source
and sensor
Reflecting object / surface
SONO/US: Ultra Sound
Components of a transducer
SONO/US: Ultra Sound
SONO/US: Ultra Sound
Different types of transducers:
Linear field
Sectoriel field
SONO/US: Ultra Sound
Different image modes:
A mode: Amplitudes
B mode: Amplitudes
converted to grey scale
M Mode: Motion curves
Doppler mode: Velocity
displayed by colors
Doppler US
(Images by S. Ohlerth, Tierspital
Zürich)
Echokardiography
(Images by S. Ohlerth, Tierspital
Zürich)
M Mode
Image Viewing and Medical Decision Making
• Suitable location:
Illumination / light!
• Equipment: special
monitors for radiology
(10 bit resolution,
calibrated)
• Good knowledge of
anatomy, physiology
and manifestations of
pathology
• Silence!
