Computed Tomography:

Physics considerations for

Quality and Dose
Martin Gunn
Outline
Frequency of CT
Bioeffects of radiation
Radiation dose in the ER
Image noise and radiation dose
kV and intravenous contrast
Shielding
Z -Overscanning
Protocol design
Prediction rules and utilization
Special considerations
USA CT Procedures / Year
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
C
T

S
c
a
n
s

(
m
i
l
l
i
o
n
s
)
Hospital Non-Hospital
Annual Growth > 10% / year. Pop growth < 1% / year
NCRP Scientific Committee 6-2, 2008
CT Scanners Per Million Population
CT Scanners /
million (OECD)
20.6 2005
15.9 1995
10.8 1990
% of ED Evaluations involving CT
Broder and Warshauer, Emergency Radiology Sept 2006 13: 25-30
CT Utilization in the ER 2000-2005
Broder and Warshauer, Emergency Radiology Sept 2006 13: 25-30
Radiation Exposure in BWH
Brigham and Womens Hospital, Boston
MA
Longitudinal study looking retrospectively
at 22 years of data.
190,712 CT exams in 31,462 patients.
Mean 6.1 CTs (54 mSv), max 132
CTs (1375 mSv).
15% > 100 mSv
4% > 250 mSv
1% > 400 mSv
Sodickson et al, American Society of Emergency Radiology Annual Meeting, Oct 2008, Houston TX
BWH Longitudinal CT Survey *
Max 1/15 (6.7%) Max 1/8 (12.5%)
Mean 1/509 (0.2%) Mean 1 / 320
(0.3%)
1% of patients >
1/62 (1.6%)
1% of patients > 1/
38 (2.6%)
3% of patients >
1/100
7% of patients > 1
/ 100
CANCER
MORTALITY
CANCER
INCIDENCE
* Sodickson A, American Society of Emergency Radiology Annual Meeting, Oct 2008, Houston TX
Diagnostic Accuracy
Increasing Utilization of CT
Diagnostic accuracy
Cx spine, appendicitis, renal colic, multi
rule out.
Replacement of other
modalities:
Volume of CT > study it replaces.
Renal CT vol. > IVU vol. for renal colic
Increased availability
Clinicans and Staff:
? Reduced tolerance for diagnostic
uncertainty or delay.
More rapid patient throughput
BEIR VII Report 2005
Supports Linear No Threshold (LNT)
Risk Model
Risk model for cancer development:
1 person / 1000 would develop cancer from
10 mSv (CT Abdomen / Pelvis)
Committee to assess health risks from exposure to low levels of ionizing radiation, National
Research Council (2005) Health risks from exposure to low levels of tadiation: BEIR VII phase
2, National Academies, Washington DC.
Relative Biological Risk of Cancer
0 50 100 250 150 200
1.01
1.02
1.03
1.04
1.05
1.06
1.00
1.07
BIER VII Report Effective Dose (mSv)
L
i
f
e
t
i
m
e

a
t
t
r
i
b
u
t
a
b
l
e

r
i
s
k

(
L
A
R
)
Linear No Threshold
(LNT)
Linear Threshold
Hormesis
The following organizations believe that the current
evidence supports the Linear No Threshold (LNT)
model of radiation induced cancer and hereditary
disease.
International Commission on Radiation Protection (IRCP)
United National Scientific Committee on Effects of Atomic
Radiation (UNSCEAR)
Radiation Protection Division of the UK Health Protection Agency
(formerly NRPB).
National Council on Radiation Protection (NCRP) (USA)
National Academy of Science (USA).
Environmental Protection Agency (USA).
International Organizations Supporting LNT Theory
Hormesis
Greek: hormaein: to excite.
Low levels of radiation exposure have a
beneficial effect, lowering the rate of
cancer compared to no exposure.
Theory is that a small radiation dose up-
regulates DNA repair mechanisms, adaptive
response.
Supported mostly by plant, protozoal and
fungal studies, a few mouse studies, and a
few human observational studies.
Nearly all studies have serious problems.
Hormesis: Position of
National Academy of Sciences (BEIR VII)
BIER VII: The assumption that any
stimulatory hormetic effects from low
doses of ionizing radiation will have a
significant health benefit to humans that
exceeds potential detrimental effects
from the radiation exposure is
unwarranted at this time.
Publicity
CT scans in children linked to cancer
By Steve Sternberg, USA TODAY, January 22, 2001
Each year, about 1.6 million children in the USA get CT
scans to the head and abdomen and about 1,500 of
those will die later in life of radiation-induced cancer,
according to research out today.
What's more, CT or computed tomography scans given
to kids are typically calibrated for adults, so children
absorb two to six times the radiation needed to produce
clear images , a second study
Evidence of Radiation Risks
Studies of humans exposed to
radiation.
Mostly from Atomic bomb survivors from
Hiroshima and Nagasaki.
Radiation Effects Research Foundation (RERF) and the
Atomic Bomb Casualty Commission (ABCC)
Insufficient statistical power at low
radiation doses (< 50-100 mSv).
Linear response above these levels.
Cellular and animal studies used for
lower levels.
Latency problem: some cancers take
20-30 years to develop.
Brenner et al, NEJM 2007 357: 2277
Lifetime Attributable Risk of Cancer Death and
Age: Abdominal CT
Brenner et al, NEJM 2007 357: 2277
Lifetime Attributable Risk of Cancer (10mGy)
Availability
Utilization.
Need for
diagnostic certainty
Concerns about
radiation
Regulation
What can we do?
1.Use technology to reduce radiation exposure.
2.Image patients appropriately
3.Track per scan and per patient radiation dose.
CT Dilemmas
CMS PQRI Test Measures 2008:
T144: COMPUTED TOMOGRAPHY (CT)
RADIATION DOSE REDUCTION
Percentage of final reports for CT
examinations performed with
documentation of use of appropriate
radiation dose reduction devices OR
manual techniques for appropriate
moderation of exposure.
CMS 2008 PQRI Test Measure Specification
http://www.cms.hhs.gov/PQRI/Downloads/PQRI2008TestMeasureSpecifications.pdf
New technologies.
Getting more for
less.
ALARA
Tube Current Modulation
Tube Current Modulation
Longitudinal Tube Current
Modulation
Angular tube current modulation
Combined (Angular-Longitudinal)
Tube Current Modulation
Cardiac CT:
ECG synchronized tube current
modulation.
Prospective cardiac gating.
Longitudinal Tube Current
Modulation
Tube Current
Varies the tube current
(mA) along the z-axis
Different mA / dose
applied to different
regions
Scout series used to
calculate mA along z-axis
to yield a pre-determined
setting for image quality
(GE = Noise Index).
0 380
Angular Tube Current Modulation
Radiation output (mA) is adjusted to
minimize dose in lower density profiles of
the patients.
Occurs during each tube rotation.
m
A
z axis of scan
T
u
b
e

c
u
r
r
e
n
t

(
m
A
)
Combined Dose Modulation
Fixed mA
D
o
s
e

t
o
o

h
i
g
h

w
i
t
h

f
i
x
e
d

m
A
Dose too low with fixed mA
Dose Savings:
1. McCullough CH Radiographics 2006; 26: 503-512
2. Hausleiter et al, Circulation. 2006;113:1305-1310
3. Shuman et al, Radiology 2008;248:431-437
Retrospective < 77%
3
Prospective
Triggering
Fixed mA < 40%
2
Retrospective
Gating
Fixed mA 0-45%
1
Combined
modulation
Compared
to:
Saving Technique
ECG Gated Tube Current Modulation
High Tube
Current
Low Tube
Current
Tube current reduced during parts of the cardiac
cycle when data not used for coronary CTA is
obtained.
Beam is on during the whole acquisition.
ECG Gated Current Modulation
m
A
Prospective ECG Triggering
Tube on Tube off
X-ray beam is on about 25% of the R-R interval.
Step and shoot technique.
Predicts timing of next R-wave.
Mean dose 6.2 mSv (2.3-11.9 mSv)
1
Shuman et al, Radiology 2008;248:431-437
Table movement
Prospective ECG Triggering
Prospective Gated CCTA
Partial Scan
Tube is turned off for part of the rotation to
avoid exposure to radiosensitive organs.
Occurs during each tube rotation.
Tube on for about 232 degrees.
Fig C: Vollmer and Kalender, Eur Radiol. 2008 Aug;18(8):1674-82
Dose Distribution
Bismuth Shielding
Z Over-scanning
Bismuth Shielding
Noise Distribution with Bismuth Shielding
Adapted from Vollmar and Kalender, Eur Radiol. 2008 Aug;18(8):1674-82
Primary beam exposure in areas above
and below the scan range.
Ends of the helix.
Wider detector arrays and higher pitches.
Overscan
Top axial slice Bottom axial slice
Z- Over-scanning
Adaptive Collimation
Top axial slice
Bottom axial slice
Collimator
Collimator
Adjusting the Scan
Parameters:
kVp / Dual Energy CT
mA,
Effective mAs or Noise Index
Reconstruction kernel.
Display window.
Reconstruction thickness
Changing the kVp and DECT
kVp: Iodine k Edge and Contrast
Attenuation of x-ray by contrast is affected
by the mean energy (keV) of the photon.
This is lower than the kVp of the beam
With increasing kVp, photon energy increases and
attenuation decreases.
At lower kVp, there is greater attenuation due to
iodine, as more photons are close to the k-edge of
I (33.2 keV)
Studies have shown an increase in contrast
enhancement of vessels (CNR) with decreasing
kVp (140 120 100 80.)
kV: Polychromatic X-ray beam
140
Photon energy (keV)
kVp
P
h
o
t
o
n

n
u
m
b
e
rk-edge of I
33.2
100 kVp 120 kVp
CTDI
vol
= 419 CTDI
vol
= 362
Same Patient, Different kVp
kVp and Dose: Exponential
80 100 120 140
0
20
40
60
80
100
R
e
l
a
t
i
v
e

C
T
D
I

(
%
)
120 kVp 140 kVp = 1.4 x in CTDI
120 kVp 80 kVp = 2.2 x in CTDI
Dual Energy CT
Dual energy scanning (typically 80
and 140kVp):
Dual source (two tubes at different
kV)
Single source with rapid kV switching.
Sandwich Detector.
Single helical acquisition.
Can generate 80kVp, 140 kVp and
virtual 120kVp, and non-
enhanced images.
Graser et al Eur Radiol. 2008 Aug 2 Epub
80 80 kVp kVp / 400 / 400 mAs mAs
140 140 kVp kVp / 96 / 96 mAs mAs
Sim Sim Un Un- -enh enh
Sim Sim 120 120 kVp kVp
Iodine Iodine Iodine + Iodine + Unenh Unenh
Radiation Dose and DE CT
No silver bullet
Dose from single DE CT ~ multiple
phase single energy (SE) CT
1,2
Single phase DECT acquisition is higher
dose than single phase SECT.
Need studies comparing image noise,
number of phases, and diagnostic
accuracy.
1.Chandarana et al, Radiology. 2008 Sep 23. Epub ahead of print.
2.Chae EJ et al, Radiology. 2008 Sep 16. Epub ahead of print.
Changing the mA or
Tube Current Modulation
Parameters.
Increased Noise Index /
Reduced Effective mAs
DLP 853
mA 439
2.5mm Recon
DLP 325
mA 244
5mm Recon
Stab wound to left flank
NI = 15.4 NI = 22.0
Reducing the mAs
Radiology 2003; 229:575580
140 kVp, 170mA, 136mAs 140 kVp, 100mA, 80mAs
CT KUB for Renal Calculi, single and 4 channel CT
scanners with fixed mA
Are Lower Dose Techniques Accurate?
AJR 2008; 191:396-401
Dose of IVU = 2.6 mSv
Low dose CT 0.7-2.1 mSv
Routinely used CT Abd Pelv = 8-16 mSv
Sensitivity = 0.966
Specificity = 0.949
Ultra-Low Dose CT Colonography
3D Colonoscopy
Optical Colonoscopy
Surgical Spec
140 kVp; and 10 mAs
Total radiation exp. (prone + supine)
1.7 mSv (M) and 2.3 mSv (F).
Optical colonoscopy:
9 Ca
2 polyps in 15 pts
Remaining 12 patients normal
Ultra-low-dose CT:
Detected all carcinomas
10 / 12 polyps (sens 83.3%).
Missed 2/6 < 5mm polyps.
Eur Rad 2003 Jun;13(6):1297-302
Noise and Windowing
WW 3000 WL550
WW 340 WL60
Bone Plus Algorithm
Reconstruction Kernel
Standard Bone Plus
Slice Reconstruction Thickness
2.5 mm
0.625 mm
Double Image Noise
Reduce Phases.
Reduce phase overlap.
Reduce scan range.
Center the patient.
Reduce Follow-up Exams.
Protocol Design
Which phases are really necessary
for multi-phase CT?
Post-contrast CT for adrenal adenomas
Non-contrast CT for HCC screening CTs
Separate dual phase vs. single phase
split bolus technique for CT IVU.
Reducing overlap between phases.
How can we position the patient to
reduce dose?
Reduce Phases: Adrenal Washout
AJR 2000;175:14111415
OR DO MRI!!
Segmented approach Segmented approach
Pan Pan- -Scan Approach Scan Approach
Overlap regions:
Overlap regions:
Wasted radiation
Wasted radiation
Reducing the Scan Range
Positioning:
Body / Profile Size and Symmetry
Noise increases with increasing
phantom diameter.
Also increases in humans, but slightly
differently, due to a number of factors
(asymmetry, tissue type, intrinsic
contrast of fat.
X-ray attenuation increases
exponentially with body diameter
Noise level doubles every 4-8 cm
increase in effective body diameter.
Asymmetric Profile
Arms at side
Arms up
Arms at side
Standard approach Standard approach
Total Body Approach Total Body Approach
mA
Patient Size
Iterative Reconstruction
Original way to reconstruct CT data.
Replaced by Filtered Back Projection
Latest statistical iterative
reconstruction techniques produce:
Less noisy images with significantly lower
radiation dose.
Less beam hardening artifact.
Currently limited by computer power.
2.5 mm
Images courtesy of GE Healthcare
Iterative Reconstruction
Quality Improvement
How to Approximate Effective Dose
eDLP Factor Region
0.019 Pelvis
0.015 Abdomen
0.017 Chest
0.0054 Neck
0.0023 Head
EUR 16262 EN-European Guidelines on Quality Criteria for Computed Tomography May 1999).
= DLP x k = 0.017
x 547.37
= 9.3 mSv
http://faculty.washington.edu/aalessio/doserisk/index.html
Repeat CT for Renal Colic
5,564 examinations performed on 4,562
patients.
61% women (mean age, 45.5 y)
38% men (mean age, 44.7 y)
3% (44) patients of pediatric age (<18 y).
Mean Eff Dose = 6.5 mSv (SDCT) & 8.5
mSv (MDCT)
176 patients (4%) had 3
examinations.
Estimated Eff Doses of 19.5 to 153.7 mSv.
All patients with multiple examinations
had a known history of nephrolithiasis.
AJR 2006; 186:1120-1124
Repeat CTs for Renal Colic*
* Does not include examinations performed at other sites
AJR 2006; 186:1120-1124
Other examinations have a proven efficacy
Quality Control
Collect Dose Data on All Scans
Effective Dose (mSV) = 0.016 x DLP
0.017 x 1710.95 = 29.08 mSv
Head 0.0023, Neck 0.0054, Chest 0.017, Abdomen 0.015, Pelvis 0.019
Other ways to reduce dose
Not doing a CT!
Ultrasound, MRI, x-ray, or no imaging.
Using prediction rules to determine
the need for imaging.
Wells criteria for CT PA for PE, New
Orleans criteria for minor head injury,
Mann-Wilson C Spine CT rules.
Summary
Increasing CT use.
Low dose CT believed to cause cancer.
New technologies can reduce dose.
Need to be vigilant in protocol design
and utilization of protocols.
Always use the dose that is reasonably
achievable.
Use prediction rules
Track dose.
Perform QA