Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
7/14/2015
1
Advances in CBCT for Breast Imaging
John M. Boone, Ph.D., FAAPM, FSBI, FACR
Professor of Radiology and Biomedical Engineering University of California Davis Medical Center
Sacramento, California 95817 Anaheim ● July 14, 2015 1
UC Davis Medical Center
2
Disclosures (required by UC Davis): CONSULTANT
Canadian Association of Radiologists
National Institutes of Health
RESEARCH FUNDING
Siemens Medical Systems
Telesecurity Systems
National Institutes of Health
ROYALTIES
Samsung Corporation
Lippincott Williams and Wilkins
TRAVEL FUNDING
Smarter Imaging Program, Canada
American Association of Physicists in Medicine (AAPM)
Nordic Association of Clinical Physicists
Japanese Society of Radiographers
International Commission on Radiation Units
California BCRP 7EB-0075 California BCRP 11I-0114 R01 CA•89260 R01 EB•002138-10 (BRP) R01 CA•129561 (RDB) P30 CA•093373 (CCSG) Susan G. Komen Foundation University of Pittsburgh UC Davis Bridge Funding California BCRP 20IB-0125 (Merced) R01 CA●181081
Acknowledgements:
2015 Annual AAPM Meeting / Anaheim / July 14, 2015 3
1. Introduction
2. Hardware & Software
3. Technical Performance
4. Patient Imaging
5. Clinical Performance
6. Summary
Advances in CBCT for Breast Imaging
7/14/2015
2
4
Mammography:
Breast CT
5 5 7/14/2015 5 5
2015 Annual AAPM Meeting / Anaheim / July 14, 2015 6
1. Introduction
2. Hardware & Software
3. Technical Performance
4. Patient Imaging
5. Clinical Performance
6. Summary
Advances in CBCT for Breast Imaging
7/14/2015
3
UC Davis Breast CT scanner development
ALBION 2004
BODEGA 2007
CAMBRIA 2012 DOHENY 2015
8
Albion
Cambria
Bodega
Doheny
UC Davis
Anaheim
9
Design Fabrication Integration
7/14/2015
4
Hardware component evolution
Albion 2003
Bodega 2007
Cambria 2012
Doheny 2015
x-ray tube
bearing
detector
640
W
1000
W
11
Doheny: Other Components
Projection Images
Pre-processing Reconstructed CT volume
Fourier Filtering
Back Projection
Felkamp Reconstruction Module
Detector Gain Calibration
Geometric Calibration
12
Calibration Software
7/14/2015
5
Detector calibration: Flat field correction
13
Raw Image
ADU(x,y) = A(x,y) + B(x,y) × mA
Gain variation between dexels ≈ 48%
calibration data files
( , ) ( , )( , )
( , ) ( , )
raw raw
corr
I x y A x yI x y g
B x y A x y
14
calibration data files
Detector calibration: Flat field correction
15
Y
X
detector plane
Geometric calibration
7/14/2015
6
16
Physical scanner geometry Reconstruction algorithm
(xc, yc) Dx
Dy
SIC
Geometric calibration
17
Feldkamp Reconstruction
18
True 3D Display !
7/14/2015
7
2015 Annual AAPM Meeting / Anaheim / July 14, 2015 19
1. Introduction
2. Hardware & Software
3. Technical Performance
4. Patient Imaging
5. Clinical Performance
6. Summary
Advances in CBCT for Breast Imaging
20
1 1 2 2 3 3( ) (1 exp( / )) (1 exp( / )) (1 exp( / ))g t f A a t T a t T a t T
Yang et al., Computer modeling of the spatial resolution properties of a dedicated breast CT system, Med Phys 2008
detector lag
gantry rotation
detector MTF focal spot size
spatial resolution modeling
21
Engineering impacts resolution
Center of FOV
Edge of FOV
Bodega
Bodega
center edge
Continuous x-ray (33 ms)
7/14/2015
8
22
Engineering impacts resolution
Center of FOV
Edge of FOV
Cambria
Cambria
pulsed acquisition (4 ms)
23
Center of FOV
Edge of FOV
Doheny
Doheny
smaller dexels
pulsed acquisition (4 ms) + higher resolution detector Engineering impacts resolution
24
total noise anatomical noise
NPS(f) = NPSa(f) + NPSq(f)
quantum noise
Detector Performance (noise)
7/14/2015
9
25
cone angle
coronal sagittal
2( , )
( , ) X Y
X Y
F u vNPS u v
N N D D
axial
Detector Performance (NPS evaluations)
26
Slice Thickness
x-ray tube mA
RECON algorithm
Shepp-Logan
RECON algorithm
Ramp
Near-Neigh
Bilinear
Yang et al., Noise power properties of a cone beam CT system for breast cancer detection, Med Phys. 2008
Noise Power Spectrum (NPS) measurements (Bodega)
2015 Annual AAPM Meeting / Anaheim / July 14, 2015 27
1. Introduction
2. Hardware & Software
3. Technical Performance
4. Patient Imaging
5. Clinical Performance
6. Summary
Advances in CBCT for Breast Imaging
7/14/2015
10
Phantom Imaging
Clinical Imaging
1. Over 600 women on UC Davis scanners
2. Suspicion of BC (BIRADS 4 or 5)
3. 16 second scan (breath hold)
4. 500 projection images acquired (1024 x 768)
5. About 200 have had contrast injection
6. Radiation dose same as 2V mammography
29
Funded for 400 more
~10 sec on new scanner
400 – 800 views on new scanner (2048 x 1536)
Developing AEC for Doheny
400 more to go, with DCE-MRI comparison also
Contralateral (normal) breasts imaged as well
30
296
January 2005
7/14/2015
11
31
bCT (no injected contrast)
32
33
7/14/2015
12
34 Pt 160
Contrasted Enhanced breast CT pre
pre post
post
Malignant
benign
35
36
Dedicated breast PET-CT imaging system
7/14/2015
13
2015 Annual AAPM Meeting / Anaheim / July 14, 2015 37
1. Introduction
2. Hardware & Software
3. Technical Performance
4. Patient Imaging
5. Clinical Performance
6. Summary
Advances in CBCT for Breast Imaging
2010
DHU AUC = 0.87
Shadi Aminololama-Shakeri, M.D.
Prospective Clinical Trial 105 patients /103 lesions (BIRADS 4 or 5) imaged on VCO mammo / tomo / CE-bCT all biopsied
2 Radiologists Rated Lesions using a 0 to 10 Conspicuity Score 0 = not seen 10 = excellent
malignant
benign
microcalcifications masses
27
18 27
31
one-way ANOVA with correction for multiple comparisons
total 45 58
7/14/2015
14
mammo tomo CE-bCT mammo tomo CE-bCT
Increased sensitivity
Increased specificity
mammo tomo CE-bCT mammo tomo CE-bCT
with radiologist analysis, Az = 0.96
b slope
41
NPS(f) = NPSa(f) + NPSq(f)
bCT, Tomo, and Mammo Comparisons
breast CT mammography tomosynthesis
N = 23 pts 1000 ROIs per
42
7/14/2015
15
PWMF Observer Performance Analysis
43
44
FT{S(x,y)}
PS(x,y) PWMF = FT-1
Pre-whitened Matched Filter
mammography
breast CT
45
Breast CT
“Mammo”
PWMF Observer
7/14/2015
16
(a) (b)
Human Observer Study: 2-Alternative Forced Choice Design
CT images projection images
46
3 radiologists / 6 physicists
47
Breast CT
“Mammo”
PWMF Observer
Radiologist Observers average of 3
2015 Annual AAPM Meeting / Anaheim / July 14, 2015 48
1. Introduction
2. Hardware & Software
3. Technical Performance
4. Patient Imaging
5. Clinical Performance
6. Summary
Advances in CBCT for Breast Imaging
7/14/2015
17
49
• Breast CT shows promise by all comparative metrics for mass lesions (beta, AUC, %correct)
• Observer performance results show enhanced mass lesion detection
• Computer observer (PWMF)
• Radiologist observers
• Breast CT (Koenig) FDA approved for diagnostic breast imaging Feb 2015
• microcalcification detection performance needs to be equivalent to mammography for bCT screening
• CE-breast CT is virtually identical as DCE-breast MRI
Advances in CBCT for Breast Imaging
Advances in CBCT for Breast Imaging
John M. Boone, Ph.D., FAAPM, FSBI, FACR
Professor of Radiology and Biomedical Engineering University of California Davis Medical Center
Sacramento, California 95817
University of California Davis
Breast Tomography Project
2015 Annual AAPM Meeting / Anaheim / July 14, 2015 50
51
7/14/2015
18
2015 Annual AAPM Meeting / Anaheim / July 14, 2015 52
1. Introduction
2. Hardware & Software
3. Technical Performance
4. Patient Imaging
5. Clinical Performance
6. Summary
Advances in CBCT for Breast Imaging
bCT, Tomo, and Mammo Comparisons
breast CT mammography tomosynthesis
N = 23 pts 1000 ROIs per
breast CT mammo
b b
Breast CT images
7/14/2015
19
axial view (~cc)
slice thickness (mm)
~10 mm
Tomo
dis
k d
iam
eter
(m
m)
tomographic angle
measured data on the breast CT system
Breast CT Images
Mammo
7/14/2015
20
Tomosynthesis Images
0 mm
55 mm
Hardware component evolution
Albion 2003
Bodega 2007
Cambria 2012
Doheny 2015
x-ray tube
bearing
detector
comet comet Varian M1500 Varian M1503
Kollmorgen Kollmorgen Yaskawa Yaskawa
Varian 4030CB TFT
Varian 4030CB TFT
Varian 4030CB TFT
Dexela 2923 CMOS