Transcript No Slide Title

AFIIM 2008 PARIS
Potential Applications of Single Source
Dual-Energy MDCT in the Diagnosis of
Cancer
Jacob Sosna, MD
Department of Radiology
Hadassah Hebrew University Medical Center, Jerusalem, Israel
Harvard Medical School, Boston, MA
• The radiological community in
Israel has been involved in CT
development since its early days
• 1975 – Elscint CT founded
• 1978 – 1st commercial scanner
• 1979 - High resolution CT
• 1992 – 1st spiral scanner (CT
Twin)
Beta sites in Israeli Radiology Departments
Haifa Plant – CT History
2007 Br iCT 128 channels, 256 slices, 0.27s
2005 Br 64-Slice
2003 Br 40-slice
2001 1st 16-slice 0.42 sec CT
1998 1st 4-slice 0.5 sec CT
1995 World’s smallest whole-body CT
1995 Gated multiphase cardiac imaging
1994 Isotropic resolution 2x0.5mm
1992 1st spiral (dual slice) scanner (CT Twin)
1986 XRT DFS technology
1979 Highest resolution (13 lp/cm)
1978 1st Elscint CT (905 scanner)
1975 Elscint CT founded
Cardiac CT Revolution
Development of
subsequent CT
scanners was
largely driven by the
prerequisites for
cardiac imaging:
Faster scans
Greater coverage
Increased spatial
and temporal
resolution
Cardiac CT
scanners
iCT
iCT
Br-64
IDT
Dual
Quad
slice
IDT
slice
16
16
Br-40 Br-64
Br-40
MDCT
WHAT IS MISSING?
Are we really good at?
• Separation between calcification and Iodine in
CTA
• Bone-removal in CTA (Cage Removal, Skull
Removal etc.)
• Soft plaque separation
• Bone mineral & bone density assessment
• Low contrast resolution (soft tissues)
• Tracking drug delivery
Are we really good at?
• Early detection of cancer
• Reliable measurements of tumors in a
reproducible way
• Still need for high dose of contrast in repetitive
manner
• Drug tracking
• Tumor vascular studies- e.g. perfusion
DUAL
ENERGY
Imaging
CT
But different elements can have same CT #…!!
CT today …It’s all about Hounsfield units ( HU or CT# )
Spectrum Decomposition Principle
Intensity
Pre-patient
Beam filtration
Low-Energy
X-ray radiation
High-Energy
X-ray radiation
KV
Dual-Energy Imaging
• Dual-energy imaging takes advantage
of differences in the degree to which
body tissues attenuate low- and highenergy photons
• These differences are used to generate
tissue-selective images
Dual Energy Imaging
• Two types of dual-energy systems
» Single-exposure system
– Two detection systems one above the other
» Dual-exposure system
– Two sequential images are obtained at 2 energy
levels with a subsecond delay between the two
exposures
– This can create misregistration artifacts due to
slight offsets in the alignment of body structures
caused by cardiac, respiratory, bowel, and
patient motion
Simultaneous Multi-Energy Detector (SMED)
X-Rays
Photons
100%
~50%
SCINT1
SCINT2
Low Energy Raw data
~50%
E1 image
+
High Energy Raw data
E2 image
----------------------------------------
=
Weighted combined Raw data
CT image
Combined
Inner
Outer
Non-enhanced
Combined
Inner
Outer
Enhanced
It’s All About the PIXEL
-106/-135
-986/1003
+23/+35
+119/147
+197/236
+329/389
+191/215
Main Pathways
The SC Benefits over conventional
64 MDCT
• Separation
• Contrast
Main Pathways
The SC Benefits over conventional
64 MDCT
• Separation
• Contrast
Preliminary Phantom Tests
140 kV, 400 mAs
Iodine
Iodine
Calcification pins
of various
concentrations
and radii
Calcium pins
~250 HU at 120 kV
Saline
2D histogram separation
Ca-I Differentiation – Calcified Aneurysm
Iodine OR Calcium Images
• Single click bone detection/ removal
• Virtual angiography
• Virtually cleansed bowel in CTC (iodine/
barium/fat)
» Major benefit in CRC screening
• Virtual Non enhanced CT
Probabilistic Separation of Iodine:
VNC Imaging
Non-Binary Probabilistic Separation
Defining a model that
generate the noisy
data (Instead of defining
separation line)
500
E1 (HU)
3 x1
0
5
2.5
450
2
1.5
For each voxel build a
Probabilistic Mixture Model
over the neighbor in the
E1-E2 plane and over
the volume neighbor.
1
E1
400
0.5
0
0
20
40
60
80
350
10
0
12
0
14
0
16
0
18
0
20
0
300
For each voxel we can
calculate the probability
to be calcium and
probability to be iodine.
E2 (HU)
250
100
150
200
E2
250
300
350
400
Probabilistic Separation
550
500
450
400
350
300
250
200
150
100
0
The color intensity is proportional to the probability
100
200
300
400
500
Purpose
To evaluate the ability to generate virtually non
enhanced CT images from enhanced clinical CT
studies and to compare image parameters to regular
non-enhanced and contrast enhanced CT
Materials and Methods
• Spectral separation
» Enhanced phase (C+)
» Virtually non enhanced CT using a probabilistic mixture
model (VNC)
» Regular non-enhanced CT (C-)
Materials and Methods
• ROI of various organs and vessels including
»
»
»
»
»
Liver
Spleen
Aorta
PV
Muscle
• 2 experienced radiologists in consensus assessed
the visibility of calcified areas in the infra-renal aorta
and artifacts
Results
• All 22 VNC studies could be obtained
• Average change from C+ to VNC
» Aorta -114.12 (SD 1.2)
» RA -135.4 (SD 16)
» PV -7.6 (SD 2.34)
» IVC -1.73 (SD 1.99)
Results
• Average change from VNC to C» Aorta 38.03 (SD 0.6)
» RA 37.8 (SD 2.8)
» PV 21.8 (SD 1.93)
» IVC 12.2 (SD 0.63)
Results
• Average change from C+ to VNC for solid
tissues
» Muscle 0.40HU (SD 0.19)
» Liver 3.5HU (SD 1.15)
» Spleen 9.2 (SD 3.15)
Results
• Average change from VNC to C- for solid
tissues
» Muscle 3.29HU (SD 0.53)
» Liver 5.79HU (SD 0.28)
» Spleen 22.2 (SD 1.72)
Results
• Total of 213 calcifications in C» 196 (92%) calcifications visualized in VNC
• “Deleted” calcifications were mainly
small areas 1-2 mm in size
• Specific artifacts
» Edge enhancement
» Metallic clips diminished
Scan without iodine
With iodine
Scan without iodine
VNC
Scan without iodine
Scan with iodine
Scan without iodine
VNC
Scan without iodine
Scan with iodine
Scan without iodine
VNC
Conclusions
• Diagnostic virtually non-enhanced images can be
obtained with single source dual energy CT
• Iodine is “deleted” mainly from vessels and does not
affect solid organs
• Majority of calcifications are preserved
• It may obviate the need for regular non-enhanced
phase in multi-phasic MDCT especially for CTA studies
Potential Applications in
Oncology
• No need for non –enhanced phase?
• Better visualization of nodules
• Benefit in adrenal imaging
Electronic cleansing: dual-energy analysis vs. HU thresholds
Intake of both Iodine and Barium
The colon is partially
filled with stool and
both Iodine contrast
and Barium contrast
Electronic cleansing
with dual-energy
analysis
Electronic cleansing
with high and low
HU thresholds only
1
Materials
1. Water
2. Calcium 0.35-0.04gr/ml diluted in water.
3. IntraLipid 20%-5% diluted in water.
4. Iodine (Meglumine Ioxitalamate ): 30 – 3mg/ml diluted in saline.
5. Gadolinium (gadoteric acid): 0.125-0.025mmole/ml diluted in
saline
6. Cis-platinume 1mg/1ml (DOTAREM- Guerbet, France).
7. Barium Sulfate suspension 0.02-0.08gr/ml diluted in saline
8. soy bean oil (92%)
9. Bovine Liver
10. Chicken breast
Results
Calcium and Gadolinium are on the same
separation line
Barium and Iodine are on the same
separation line
We are not sensitive to lipids concentrations
<20%(shall be further tested)
Cis platinum not detected reliabley
Further interest in drug detection with specific
absorption patterns
CT numbers
5. Calcium
Avg 306
6. Gadolinium
Avg 362
7. Cis Platinum
Avg 26.6
4. Barium
Avg 488
3. 20% oil
Avg: -16
8. Water
Avg 1.3
2, Oil
Avg: -102
1. Iodine
Avg: 319
Materials Separation
5. Calcium
4. Barium
6. Gadolinium
3. 20% oil
7. Cis Platinum
2, Oil
8. Water
1. Iodine
Main Pathways
The SC Benefits over conventional
64 MDCT
• Separation
• Contrast
Attenuation of different materials
2
10
1
0.06
0.05
Attenuation
10
Iodine shows higher
signal
Iodine shows lower
signal
0
10
Calcium
Calcium
K
0.04
-1
10
0.03
Iodine solution
Iodine solution
Tube
0.02
0.01
-2
0
10
0
20
20
40
40
60
60
80
80
100
100
120
120
140
kev
reduction in tube voltage leads to an increase in attenuation of
iodinated contrast material
Iodine Augmentation in vivo
• 18 patients 61 measurements (CTA)
» Average increase of 33.4% in density in CTA
• 30 patients 81 measurements (non CTA)
» Average increase of 32.0% in density in cases
CTA Results
Regular Low
Image energy
3 cc
228
320
5 cc
324
471
7 cc
483
720
28.5% contrast dose reduction
Regular
Low
Low dose
Regular
Low
Medium dose
Regular
Low
Medium dose
Regular
High dose
Low
Human Applications
• In adults 33% augmentation in density
observed with low energy imaging
• The difference may be explained by
smaller size of rabbits (3.5 kg)
• Expected similar change in density in low
energy imaging in infants
» Not performed yet due to IRB regulations
Application in Oncology Patients
• Significant iodine density augmentation can be
obtained using the low energy layer versus the
clinical image
• A reduction in contrast volume may be feasible
with similar densities at CT studies
Contrast
• Oncological imaging
–Lesion conspicuity (liver,
kidney, breast)
–Improved CAD
The Future in Oncology Imaging
Abdominal CT
DE Abdominal CT
C-
needed
Not needed
C+
needed
Needed but with
reduced contrast
Drug tracking
Not possible
possible
Lesion detection
good
improved
6 '‫מלכים ב' טז‬
".‫ עַ ד הַ יֹום הַ זֶּה‬,‫ וַ ֵישְׁ בּו שָּ ם‬,‫ בָּאּו אֵ ילַ ת‬....."
Hosting AFIIM
Hotel Dan Eilat (on the Red Sea)
29-31 October 2008
Lectures and presentations in English