Transcript MARS_talk_MAH_CERN - University of Canterbury

The MARS Project
Current Status
Future Directions
Dr Mike Hurrell
Consultant Radiologist
No financial interests to
declare
Talk Outline
►The MARS project in brief
►Recent achievements
● Liver fat
● 3D visualisation
MARS Project Goal
To develop clinical applications of
spectral CT using the Medipix chip
What is MARS?
►A scanner: Medipix All Resolution System
►A collaboration
● International
 CERN, Geneva (Medipix chip)
 Erlangen University, Germany (Small animal lab)
 Mayo Clinic, USA (Small animal lab)
● New Zealand
 University of Canterbury (Computing, Physics, Engineering
Workshops)
 University of Otago (Christchurch School of Medicine)
 Canterbury District Health Board (Radiology Department)
Useful Medipix chip features
►65,536 spectrum analysers on a 14mm chip
● Each 0.055mm x 0.055mm
►Very low noise
►Xray stopping power equivalent to modern CT
detectors
● Choice of detector layers (Si, CdTe, GaAs…)
►Fast readout (kHz)
MARS scanner (Model 2)
MARS scanner (Model 2)
Medipix Detector
Sample
X-ray tube head
K-edge and attenuation
Attenuation Spectrum
Photoelectric Effect
I Ba
L
M
K
Gd
Dual Energy Radiographs
Dual Energy CT
Iodinated contrast vs Bone
Johnson, Krauß, Sedlmair et al. Material differentiation by dual energy CT: initial
experience. European Rad (2007) 17:1510
CT pseudocolor failure
– vessel discontinuities
CT pseudocolor failure
– mapping misrepresentation
3D volume rendered
CT slice
Scout
Medipix3 / MARS
vs Dual Energy CT
►
►
►
Multiple (up to 8) settable energy e.g. >70 >80 >90 >100keV
limits
93keV photon will be counted in detectors #1, #2 and #3
All photons above lowest threshold
are used
Shorter exposure reduces movement
artefact and dose
●
►
►
No double scanning at second energy
Non-overlapping energy bins gives
better characterisation
Software adjustable energy bins to
optimise discrimination
●
●
match target material
reject confounding material
Siemens
Energy information in CT
Standard CT
Hounsfield
Units
Grey
scale
Xray source
Dual energy CT
Xray source
Grey
scale
Xray source
Grey
scale
Calcium
or Iodine
or…
MARS-CT
Medipix
(colour)
Xray source
Source
Calcium
+ Iodine
+ Barium
+ Gad
+…
Voxel
Detector
Result
Recent Projects
►Fatty liver
►(Selective imaging of contrast agents)
►3D visualisation
Recent Projects
►Fatty liver
►(Selective imaging of contrast agents)
►3D visualisation
Fatty liver quantification
Kyra Berg, Mike Clark, James Carr
L
L
S
S
Normal
Fatty liver
Fatty liver quantification
►Aim
● Generation of linear absorption curves
● Measure fat concentration (gm/cm3)
►“Low threshold only” mode, 7x keV settings
● 12.5 15.0 17.5 20.0 25.0 30.0 35.0 keV
Attenuation Curves
1.6
Liver
Fat
Blend
Liver
Fat
Blend
Linear Attenuation (A/cm)
1.4
1.2
1
0.8
0.6
0.4
0.2
0
12.5
17.5
22.5
Bin Energy (keV)
27.5
32.5
Summer Project Summary
►Established proof of concept
►Issues regarding
● Medipix chip reliability
 Radiation damage to electronics (Si >> CdTe sensor)
 Calibration stability
 MUROS board power supply
● Reconstruction normalisation
 Octopus application
 Prevented reliable pixel measurement
Recent Projects
►Fatty liver
►(Selective imaging of contrast agents)
►3D visualisation
Feasibility of selective imaging of contrast
agents
Anthony Butler, Mike Clerk, Marcus Firsching
C-
C+
Delayed C+
PCA/Multiple materials
iodine
calcium
air
Triple Phase Protocol
► CT
► MARS

C- scan

Contrast1 outside room

Contrast and scan

Contrast2 on scanner

Delay and scan

Scan x1

3 scans

1 scan

Twice on table

Once on table
Recent Projects
►Fatty liver
►(Selective imaging of contrast agents)
►3D Viewer
Multispectral CT:
3D viewer
Niels de Ruiter
Mouse12
>23keV Green
k-edges: Calcium=4.0
>30keV Blue
>35keV Red
Iodine=33.2 keV Barium=37.5 keV
PCA
3 data channels
Summer Project Summary
►Successful implementation of 3D viewer
►LAC: slope and K-edge
● AbsorptionkeV → colour
 Colour tint effect
● PCA → colour
 Better characterisation (Barium/Iodine)
Future work
►Currently active projects
● Atherosclerotic plaque characterisation in vitro
 Predict instability and monitor treatment effect
● Breast imaging
 Surgical specimens
● Liver/fat characterisation
 Lipid quantification
►Medium term
● Multi-contrast CT (“unenhanced” + Gd + I + Ca)
 PCA (continuation)
● Platinum: cis-platin treatment for malignancy
 Platinum 78.4 keV K-edge
 Tumour distribution and measure focal recurrence
● Antibody-labelled nanoparticles/liposomes (cellular targets)
● Beam hardening
Conclusion
Recent achievements
►“Proof of concept” characterisation
● Fat vs Liver
►3D Viewer
● “Proof of concept” colour display
● Contrast agent separation
►MARS component upgrades
● Hardware development
 New chip (Medipix3)
 New Medipix interface board
 New CT scanner
● Software development
 Cone-beam reconstruction (beta)
 Data structure (Deep Server)