Transcript Group 1C Goals (cross

QIBA CT Volumetrics Cross-Platform Study
(Group 1C)
March 29, 2009
Interclinic Comparison of CT Volumetry
Quantitative Imaging Biomarker Alliance
Charge
1.
2.
3.
4.
5.
To agree on the scanner settings and other
protocol elements under which imagery is to
be collected.
To agree on requirements of phantoms to be
imaged/measured.
To agree on the platforms and centers to be
selected for imagery collected.
To identify the measurements and the
algorithms for use in image processing.
To specify the analysis of the measurements.
Goals
1.
Measure the volume of nodules on CT imagery
collected from several CT scanners and sites
(may include multiple settings on single
scanners).
1.
Measure image noise and other image quality
factors and determine their impact on the
measurement of volume.
2.
Compare the accuracy and precision of volume
measurements for these phantom datasets.
3.
Determine the minimum detectable level of
change that can be achieved when measuring
nodules in phantom datasets.
Goal 1
1.
Measure nodule volume on CT imagery
collected from several CT scanners/sites
(including single scanners with varying
settings). Determine the systems to be used
and the system settings to be varied.
(a)
(b)
(c)
(d)
(e)
kVp may be specified.
mAs may be specified.
collimation fixed (+)
field of view (skin-to-skin = closest possible view)reconstruction filters – follow-up Wendy & radiologists
- Find “equivalent” filters.
Site selection – poll the team for potential
image collection sites.
Goal 2
2.
Measure “image noise” and determine its
impact on the measurement of volume.
Facilitates inter-comparison of scanner
results.
(a)
Characterizing / specifying image
quality
(a)
Expert visual assessment of image
quality. (Deni)
QIBA CT 1-C Protocol
- first branch
Using protocol for NLST (ACRIN 6678?)
specify kVp, slice thickness, mAs, rotation
time, pitch, reconstruction kernel (affects
MTF). Use (water and ACR) phantoms to
characterize the resolution and noise levels
under this branch of the protocol.
Sample Protocol Chart for NLST
Cagnon et al.
Parameter
Siemens
Vol Zoom/
Sensation 4
4-slice/0.5 sec
4 x 2.5
Siemens
Vol Zoom/ Sensation 4
4-slice/0.5 sec
4x1
Siemens
Sensation 16
16 x .75
Siemens
Sensation 64
64x.0.6
(beam collimation 32x0.6)
120
120
120
120
Gantry Rotation Time
0.5 sec
0.5 sec
0.5 sec
0.50 sec
mA (Regular patient-Large patient values)
75-150
80-160
75-150
50-100
mAs (Reg-Lg) 1
37.5-75
40-80
37.5-75
25-50
Scanner effective mAs2 (Reg-Lg)
25-50
20-40
25-50
25-50
Detector Collimation (mm) - T
2.5 mm
1 mm
0.75 mm
0.6 mm
Number of active channels - N
4
4
16
32
Detector Configuration - N x T
4 x 2.5 mm
4 x 1 mm
16 x .75 mm
32 x 0.6 mm
N/A
N/A
N/A
64x0.6mm
15 mm
8 mm
18 mm
19.2 mm
1.5
2
1.5
1.0
Table Speed (mm/second)
30 mm/sec
16 mm/sec
36 mm/sec
38.4 mm/sec
Scan Time (40 cm thorax)
13 sec
25 sec
11 sec
11 sec
Nominal Reconstructed Slice Width
3 mm
2 mm
2 mm
2 mm
2.0 mm
1.8 mm
1.8 mm
1.8 mm
Reconstruction Algorithm3
B30
B30
B30
B30
# Images/Data set (40 cm thorax)
200
223
223
223
kV
Collimation (on operator console)
Table incrementation (mm/rotation) - I
Pitch ([mm/rotation] /beam collimation) I/NT
Reconstruction Interval 3
4
QIBA CT 1-C Profiles
- second branch
Specify PERFORMANCE metrics such as simple
spatial resolution and noise metrics.
 kVp
(affects contrast difference between materials)
 Slice thickness, recon interval (affects z-axis resolution &
noise)
 Rotation time and pitch (coverage, breath hold, etc.)
 Recon kernel OR recon kernel performance - EXAMPLE:
– Choose kernel such that you can see 6 or 7 (but no more
than 7) lp/cm on ACR phantom….or
– 10% MTF should be between 6 and 7 lp/cm
 mA
level performance
– Choose effective mAs level so that std dev is between 20
and 30 HU in a 20 cm water phantom
Goal 3
Compare the accuracy and precision of
radiologists’ measurements of
RECIST and Volume for these
phantom datasets. (image mask?)
a)
b)
c)
d)
RECIST vs. volume.
Investigate variance & bias.
Inter-system variation.
Intra-system variation.
Goal 4
4. DEFER: Determine the minimum
detectable level of change that can be
achieved when measuring nodules in
phantom datasets.
Required resources



Use of FDA phantom, water and
ACR phantoms.
Recruit clinical image collection sites
through QIBA-CT group.
Use QIBA CT 1-A mark-up at RadPharm, generating


RECIST
Segmented volume.
Future steps
Refine Questions and Experimental Design.
 Recruit participating clinics. Share and
discuss plans with associated medical
physicists.
 Confirm availability and discuss reading with
Rad-Pharm.
 Confirm availability and schedule FDA
phantom.
