Latest Advances in Digital Breast
Tomosynthesis
Wei Zhao, Bo Zhao and Yue-houng Hu
Dept. of Radiology
State University of New York at Stony Brook
Breast Tomosynthesis
Limitation of Mammography
• False diagnoses due to superposition of
breast tissue
• Solution: 3D imaging
– Breast CT: Cone-beam CT (180°) with
dedicated prone table
– Breast tomosynthesis: modified mammo
gantry with limited angular range (15o-60o)
Mammogram
Tomosynthesis
• Fast clinical transition
• Limited angular range:
15~60o
• Slice thickness: 1mm
• Projections: 11- 50
• Dose: ~1 - 2 times
single mammogram
Biopsy proven cancer
Courtesy: B. Ren, Hologic
Different DBT Design
Company
System param.
Views
Detector
Scan time
(s)
Recon.
GE
±20°, step/shoot
(MGH, Sunnybrook)
15
CsI/a-Si
100 um
15-23
MLEM
SART
±30°, step/shoot
(U. of Michigan)
21
7s
SART
Hologic
±7.5, continuous
Multiple sites
11
a-Se, 70 um
2x2 binning
10
FBP
Siemens
±22°, step/shoot
Duke,Malmo,SUNY
25
a-Se, 85 um
12.5/20
Bin/full
FBP
Dexela
±12-20°
(U. of Virginia)
13
Fiber optic
coupled CCD
30 s
MLEM
X-counter
±13°
48
Gas counting, 48
slit, 60 um
Sectra
±5.5º, 6-site trial
21
Si counting, 21
slit, 50 um
FBP
iterative
3-8 s
Factors Affecting Image Quality
•
Detector performance
– DQE at low dose: 1/NView
– Temporal performance: lag and ghosting
• Imaging geometry
– Angular range, number of views
– Focal spot blur: continuous tube travel
• Reconstruction algorithm
– Analytical: FBP (filter design)
– Iterative
Imaging Geometry: spatial vs.
frequency domain
Detector Performance
0.8
data-line direction
nbin bin
0.7
fz
0.43 mR
2.15 mR
6.01 mR
0.6
0.5
DQE
dz
Reconstructed
slices
dx d y
voxel
space c overed by
ac quired im ages
rec tangular area (fzNY,fx NY)
used for reconstruction
0.4
θ
fz
NY
fx
z
0.3
fx
y
NY
0.2
x
0.1
0.0
0
1
2
3
4
spatial frequency (cycles/mm)
5
6
Frequency domain
B. Zhao and W. Zhao, Med. Phys. 2008
Spatial domain
3D Image quality metrics:
MTF, NPS and DQE
Noise Measurement
Focal spot
fz
fz
4cm Lucite
2fNY
2fNY
Aliasing
dz
dx
voxel
fNY
fx
Reconstructed
3d image
fx
Θ
z
40
slices
dy
3D FFT
y
x
Slice thickness 1 mm:
Noise aliasing in z
Projection
noise image
Slice thickness filter:
eliminate aliasing
Voxel dimension
dx=0.085 mm
dy=0.085 mm
dz=1 mm
detector
B. Zhao et. al, Med. Phys. 2009
NPS vs. angular dose distribution
In-depth NPS
fz
measurement
fy
experiment
model
fx
ADS-1
SBP
ADS-2
Ramp x Hann
(
2fNY
Ramp x Hann x
ST (width
=0.47cycles/mm)
fz
ADS-3
model
ADS-1
fx
•Effect of angular range
•Effect of reconstruction filters
•Effect of 3D sampling, ST filter
ADS-2
fz
fx
ADS-3
3D NPS
In-plane NPS
avg NPS
NPS of top slice
NPS from model
60000
Projection 2D NPS
fy
fx
50000
calculated 2D avg
NPS
2
40000
2
Measured 2D avg
NPS from all slices
NPS(ADU mm )
Measured 2D NPS
from top slice
30000
20000
fy
10000
fx
0
•Reconstruction: Ramp + Hanning
-6
-4
-2
0
2
4
6
freq(cycles/mm)
Oversampled Point Spread Function
(PSF)
PSF/artifact vs. Angular Range
x (mm)
y
z
z
0
x
y
70 um wire tilted 20 degrees
17
(mm)
0
14
+20o
+15o
+10o
+5o
MTF – Dependence on Filters
Impact of Angular Range on
In-Plane MTF
x (cyc/mm)
SBP
3.0
angular range
o
10
o
20
o
40
o
60
o
180
2.5
z
RA
MTF
2.0
RA+SA
1.5
1.0
0.5
0
RA+SA+ST
0.0
0
2
4
freq(cycles/mm)
0.5 cyc/mm
0
10
• Increasing angular range improves MTF at low frequencies
In-Plane MTF
In-Plane MTF Measurement
Projection 2D MTF
1.0
Focal spot
4000
fy
3500
Measurement
Model
recon setting: Rampx Hann xST, prebin
0.8
3000
fx
2000
1500
In-plane MTF
1000
0.6
MTF
ESF
2500
Reconstructed
in-plane im age
0.4
500
0
20
40
60
80
100
120
140
160
180
200
position(pixel)
0.2
Derivative(LSF)
0.0
0
2
4
6
8
freq(cycles/mm)
Edge phantom
|1D FFT|: MTFx
2cm lucite
Low f drop caused by ramp filter and limited angular range
detector
Edge phantom: 0.2mm Al
High f drop caused by other reconstruction filters
10
Dependence on Angular Range
Reconstruction filter designs
2.0
• Ramp + limited angle:
loss of low frequency
(breast density)
• Modified filters:
recovers density, but
increased out-of-plane
blur
• Iterative recon:
improvement at the
cost of computation
±10o
SDNR=2.5
SDNR=1.4
ramp
1.5
amplitude
±20o
Slice thickness (fz)
1.0
Polynomial,
similar to SART
apodization
0.5
0.0
0
1
2
3
4
Spatial frequency (cycles/mm)
ACR phantom imaged at 28kVp, W/Rh, 1.7mGy
J. Zhou et. al. Med. Phys. 2007; Ludwig et. al. IWDM2008; B. Ren, et. al. SPIE 2009
Recon: FBP with ST filter
Reconstruction filter comparison
Clinical Application of DBT
• Screening/diagnosis
• Contrast enhancement: angiogenesis
• Multimodality imaging:
– DBT + automated ultrasound (AUS)
• P. Carson et. al. U. Michigan
– DBT + optical tomography
• Boas et. al. MGH
– DBT + limited angle SPECT
FBP with Ramp
SIRT
Adapted from Ludwig et. al. (IWDM 2008)
Polynomial
• M. Williams et. al., U. Virginia
5
6
Dynamic Contrast-Enhanced (CE)
Tomosynthesis
p19_scout_enhan3
p19_scout_enhan3
Pre
X-ray projections
CE-Tomosynthesis
Post
Post
Pre
-
Tomo
reconstruction
=
Mammo
2
Signal
mg/cm
Tomo
CE-Tomo
Kinetics
3
2
1
0
0
2
4
6
8
Time (minutes)
Subtraction (CE)
Courtesy Dr. R. Jong
Courtesy M. Hill and J. Mainprize
Case 1: poorly differentiated invasive ductal carcinoma
Temporal vs. Dual-Energy CE-DBT
•
DS DBT system, GE Medical Systems
•
Automated tube motion
•
No anti-scatter grid
•
HE:
– Rh target - 0.25 mm Rh + 0.25 mm
Cu filter
– 49 kVp
•
LE
– Rh target – 0.25 mm Rh filter
– 32 kV
•
7 projections: 40° arc, 6.7° apart
•
Dose per HE/LE data set ~single
mammographic view (~2 mGy)
•
Filtered backprojection
Courtesy Ann-Katherine Carton
Courtesy
Ann-Katherine
Carton
Carton
A.-K., Chen. et al., Br. J. Rad.
2010
U. Michigan
Automated US Scanning System
GE Gen II
Research
Tomo Unit
60o angle
21 projections
7.5 sec
GE Logiq 9
US system
US
transducer
xy
translator
Dual-modality
compression
paddle
TPX
paddle
• Translator & transducer are flipped up out of field for
tomo acquisition
Stationary
digital x-ray
detector
• Translator & transducer flipped down for AUS acquisition
Courtesy: Drs. P. Carson and M. Goodsitt
Courtesy: Drs. P. Carson and M. Goodsitt
Co-location Reader Study Display GUI
showing Corresponding VOI’s in DBT & AUS
Invasive
Cancer
mammo
•
tomo
AUS-suggested ductal
extension
Courtesy: Drs. Carson & Goodsitt
AUS
New detectors: Improve low dose
performance
Further Development for DBT
Systems
x-rays
x-rays
• Gantry:
HV bias
electrode
O
L
– Faster scan, more views
O
N
T
R
Structured
CsI
MU LT
IPLE
IN
G
-
C
-
R
O
L
T
O
N
N
C
S
XER
G
D
S
C
A
N
IN
G
N
A
N
– dual-energy, shorter pulse/less focal spot
blur
High gain
photoconductor
S
C
• Generator/tube:
storage
capacitor
Avalanche
TFT
a-Se (HARP)
pixel
electrode
S
MULT
G
D
IP LE
XE R
storage
capacitor
TFT
pixel
electrode
• Detector:
– Better low dose performance, faster readout
– photon-counting
Summary
• Physics of DBT: 2D 3D
– Optimization of acquisition and recon.
• New clinical applications:
– Contrast enhanced CEDBT
– Multimodality imaging
• Challenges for further system
development
Direct FPI:
Photoconductor with
higher gain
Indirect FPI:
Avalanche photodetector
Acknowledgements
• Financial support:
– Siemens AG
– Army Breast Cancer Research Program
– NIH
• Contributing slides:
– Drs. Baorui Ren, James Mainprize, Mitch
Goodsitt, Ann-Katherine Carton, Mats
Danielsson