Department of Radiation Oncology

1
From Anatomoic Image Guided to Biological Image
Guided Radiation Therapy
Lei Xing , Ph.D.
Department of Radiation Oncology Stanford
University School of Medicine
2

• Current Research Projects
• 4D CT, CBCT, and PET image reconstruction
• Scatter, noise and motion artifatcs removal in
  4D imaging
• Real-time monitoring of tumor motion during
  dose delivery
• Inverse planning strategy for modulated arc
  therapy, 4D RT and,
• Adaptive radiation therapy
• Biological imaging, modeling and biological
  image guided radiation therapy
• Nanotechnology for radiation oncology

3
IGRT Roadmap
Pt setup and treatment delivery
3D modeling
Treatment planning
Imaging
3D/4D CBCT
4D imaging Biological imaging
4D planning
4D modeling
Adaptive therapy (imaging, planning, delivery)
Gated tx planning
Day 0
Day 14
Day 24
4
Targeting in current radiation oncology
Intra-fraction organ movement, in particularly,
respiratory motion
Inter-fraction organ movement
Target volume definition localization
5
IGART Roadmap
Pt setup and treatment delivery
3D modeling
Treatment planning
Imaging
3D/4D CBCT
4D imaging Biological imaging
4D planning
kV/MV
4D modeling
Adaptive therapy (imaging, planning, delivery)
Gated tx planning
Day 0
Day 14
Day 24
6
4D CT
t5 sec
t0 sec
7
Dynamic or 4D PET
8
4D PET Imaging
Phantom Results
3D PET
Motion direction
4D PET (after post-acquisition data processing
using our new algorithm)
Motion direction
Intensity profiles along the motion direction
B. Thorndyke, E. Schreibmann, A. Koogn, and L.
Xing, Med. Phys. 2006
9
B. Thorndyke, E. Schreibmann, A. Koogn, and L.
Xing, Med. Phys. 2006
A Liver Cancer patient
GE 4D PET
Conventional 3D PET
New 4D PET technique
1 cm
3D PET --- the lesion in the ungated image is
elongated, and mislocalized superiorly by 1
cm. GET PET location is right but signal is
week. RS 4D PET location and signal are great?.
3D PET give wrong location and wrong volume
10
Model-based Reconstruction

• Central idea

(
c
)
(
a
)
(
b
)
stationary object
deformation
virtual path
T. Li, B. Thorndyke, E. Schreibmann, Y. Yang, and
L. Xing, Med. Phys. 33, 1288-98, 2006
11
Calculated TUV for a lung tumor with and without
motion correction
12
Cone Beam CT Radiation dose, scatter, noise,
motion artifacts
Linear accelerator with onboard cone-beam CT
movies
13
Cone beam image reconstruction
Varian Medical Systems
14
Scatter
Single Row of Detectors
Transaxial
15
Scatter
Multiple CT Detectors
16
Scatter
Cone-Beam CT
17
Scatter Artifacts

• Large Scatter-to-Primary Ratios (SPR) in CBCT
  cause severe cupping/shading artifacts.

Wide collimator, high scatter
narrow collimator, low scatter
Display window min max no anti-scatter grid,
no scatter correction.
L. Zhu, J. Wang and L. Xing, Med. Phys. 2008
18
Work Flow
CBCT Projection
Registered scatter estimate
Subtract
Reconstruction
Reconstruction
Rigid registration
Partial CBCT projection
Reconstruction
Scatter corrected CT image
Scatter estimation using interpolation
Scatter estimate
L. Zhu, J. Wang and L. Xing, Med. Phys. 2008
19
Scatter noise in post-processing methods
Measurement-based scatter correction, PWLS noise
suppression, (Wang et al., 2006) Noise in the
ROI 9.75e-7
No scatter correction, no noise suppression,
Noise in the ROI 1.01e-6
Measurement-based scatter correction, no noise
suppression, Noise in the ROI 1.01e-5
L. Zhu, J. Wang and L. Xing, Med. Phys. 2008
20
Ultra-low dose CBCT
(a) (b)

80mA
10mA
10mA
J. Wang L Xing, PMB, 2008
21
Ultra-low dose fluoroscopic imaging
(a)
22
Motion artifacts in fan beam CT and CBCT
23
Motion artifacts in fan beam CT and CBCT
Static phantom



Motion phantom CB CT
Motion phantom fan beam CT
T. Li, E. Schreibmann, Y. Yang, L. Xing, PMB 2005
24
CBCT vs conventional CT for a moving phantom
 

 



Phantom study for the influence of motion in CBCT
imaging. The top row shows the CT image, and
bottom row shows the CBCT image of the same
phantom. Left column contains images of the
phantom without movement right column contains
images of the same phantom moving laterally with
a period of 4 sec and amplitude 1 cm. Serious
distortion and artifacts were produced by the
motion in CBCT image.



25
4D CBCT using Trilogy
26
CBCT projections before and after phase sorting
Stanford Research
27
CBCT phantom images
Static phantom - 3D CBCT
motion switched on - 3D CBCT
motion switched on - 4D CBCT
Li, Koong, Loo, Xing, Med. Phys., 2006
28

• Slow Gantry Rotation

29
4D CBCT 4D CT
Li, Koong, Loo, Xing, Med. Phys., 2006
30
IGART Roadmap
Pt setup and treatment delivery
3D modeling
Treatment planning
Imaging
3D/4D CBCT
4D imaging Biological imaging
4D planning
kV/MV
4D modeling
Adaptive therapy (imaging, planning, delivery)
Gated tx planning
Day 0
Day 14
Day 24
31
Deformable Image Registration for IGRT
to establish point-by-point correspondence
between two images
y
Original location
Displacement of a voxel
New location
x
To be determined
32
Beauty and Beast Transformation
Deformable Image Registration for IGRT
y
x
33
Segmented Deformable Image Registration For
Improved Modeling of the Shear Movement of the
Lungs
34
(No Transcript)
35
3.5D Radiation Therapy
4D Patient Model
- Better target delineation
36
Y. Xie, P. Lee, L. Xing, Med Phys, 2008, in
press.
37
Contour propagation
M Chao, E. Schreibmann, T. Li, L. Xing, IJROBP,
2007

Y. Xie, M. Chao, P. Lee, and L. Xing Med Phys,
2008
38

• Current Research Projects
• 4D CT, CBCT, and PET image reconstruction
• Scatter, noise and motion artifatcs removal in
  4D imaging
• Real-time monitoring of tumor motion during
  dose delivery
• Inverse planning strategy for modulated arc
  therapy, 4D RT and,
• Adaptive radiation therapy
• Biological imaging, modeling and biological
  image guided radiation therapy
• Nanotechnology for radiation oncology

39
3.5D 4D Treatment Planning
Adapted from Y. Yang
40
(No Transcript)
41
Radiation Therapy Chain Process
Real-time information of tumor position
42
(No Transcript)
43
Brainlab/Varian/CyberKnife
44
Simultaneous kV/MV imaging guided RT
delivery (R. WiersmaL. Xing, Med. Phys., 2008)
45
(No Transcript)
46
Scheme of setup verification with fiducial markers
Linear Accelerator
Treatment at same position
EPID
Matching of markers to reference image of CT
(image registration)
Vision
Adjust the setup of patient until satisfactory
position(lt 2 mm) achieved
47
(No Transcript)
48
(No Transcript)
49
(No Transcript)
50
Tracking a stable/moving fiducial during an arc
delivery
W. Liu et al, Med Phys. Submitted.
RMS (mm) 0.08 (LR), 0.07 (SI), 0.11 (AP), 0.16
(mag) Range 0.49 mm
51
Fiducial Detection Example
Match filter based fiducial detection algorithm
52

53
(No Transcript)
54
Intra-fractional prostate motion
Y. Xie, D. Dajaputra, C. King, L. Xing, IJROBP,
2008.
55
Courtesy of M. Sharpe
50min Sagittal Cine-MR
Full Rectum Empty Rectum
56
Prostate Patient (S 42) 2D BEV Tracking
Track fiducials by MV beam
30 20 10 0 -10 -20 -30
U (mm)
30 20 10 0 -10 -20 -30
V (mm)
0 10 20 30
40 50 60
70
Beam On Time (seconds)
57
IMRT treatment fields
58
(No Transcript)
59
(No Transcript)
60

• Current Research Projects
• 4D CT, CBCT, and PET image reconstruction
• Scatter, noise and motion artifatcs removal in
  4D imaging
• Real-time monitoring of tumor motion during
  dose delivery
• Inverse planning strategy for modulated arc
  therapy, 4D RT and,
• Adaptive radiation therapy
• Biological imaging, modeling and biological
  image guided radiation therapy
• Nanotechnology for radiation oncology

61
Using total-variation for IMRT inverse treatment
planning
L Zhu L. Xing, PMB, in press.
62
Inverse planning for modulated arc therapy with
incorporation of prior angular knowledge
63
3.5D 4D Treatment Planning
Adapted from Y. Yang
64
4D Adaptive Radiation Therapy
4D simulation
4D Plan
Dose Matrices
Overall DVHs
4D Verification Delivery
65
4D RT Treatment Plan




Optimize dose distribution in spatial and
temporal domains
Y. Yang, S. Huq, L Xing, Med. Phys, 2006
66
Adaptive Therapy
Simulation
Delivery
Pt setup
Planning
67
IMMOBILIZATION DOES NOT ALWAYS WORK!
68
IMMOBILIZATION DOES NOT ALWAYS WORK!
CBCT imaging of a rectal cancer patient during a
course of RT
1st wk (planning CT)
2 wk
overlay
4 wk
P. Lee, K. Goodman, L. Xing, et al, 2006 ASTRO
69
RTCT
CBCT-2
Re-panning or not? The dosimetric difference can
be up to 3-8 Gy.
70
Conventional RT vs Adaptive RT
71
(No Transcript)
72
Adaptive Radiation Therapy

• What are needed to bring ART into clinic?

• CBCT.
• Deformable model.
• Automated contour mapping from pCT to CBCT.
• Retrospective dose reconstruction.
• Deformable registration for cumulative dose
  calculation
• Inverse planning for ART
• Dose shaping tool.

73
MLC log-file generated Fluence Map
MLC log-file
MLC Workstation

• every 50 ms
• leaf position beam status

in-house program
TPS
Delivered fluence map
Actual leaf sequences
Delivered dose distribution
74
Figure 2.10 (a)
75
Calculated Vs Delivered Fluence
L. Xing J. Li, Med. Phys. 2002
76
Serial CBCTs Procedure

• CBCT taken at corrected set-up
• 3 CBCTs during the course of Tx

• interval balanced amongst the radiation dose,
  workload, and the process of the anatomical change

77
(a)
Figure 2.9
78
Use CBCT for Dose Verification
Planned (IMRT)
DVHs (planned vs delivered)
Prostate
PTV
Delivered (reconstructed dose on CBCT)
Prostate
PTV
79
MLC Workstation

• every 50 ms
• leaf positions
• gantry angle logged

in-house program
?
Treatment Planning System
DCAT Dose Reconstruction
Regenerated Leaf Sequence File
80
Gantry Angle (deg.)
Leaf 30A
Leaf 30B
Leaf position (cm)
81
(a)
(b)
axial
axial
coronal
coronal
sagittal
sagittal
82
Fig. 5
(b)
(a)
axial
coronal
sagittal
83
Original plan


To be delivered without adaptation
With adaptation
Q. Wu et al, Phys. Med. Biol. 53, 673691, 2008.
84
 
Figure 2.14. (b)
85
PET/CT
IGRT Tomorrow Molecular Imaging, tumor target
definition Biological conformal radiation
therapy (BCRT)
Where is the tumor?
86
CT
PET
CT
PET
87
The Current Imaging Toolbox
Method Minimum DetectableSize (f) Minimum Detected Cells (n)
CT 12 mm 400,000
MRI 12 mm 400,000
MRSI 7 mm (3mm at 3T) 1,000,000
SPECT 46 mm 600,000
PET 35 mm 400,000
HFUS lt1 mm 100,000
Optics 0.02 mm 1000
88
Biologically Conformal Radiation Therapy (BCRT)
Spatial distribution of biological parameters
Biological imaging
Spatially non-uniform conformal dose distribution
89
Prescription for molecular/functional image
guided IMRT
Yang Y and Xing L, Med. Phys. 32, 1473-84, 2005.
90

• Current Research Projects
• 4D CT, CBCT, and PET image reconstruction
• Scatter, noise and motion artifatcs removal in
  4D imaging
• Real-time monitoring of tumor motion during
  dose delivery
• Inverse planning strategy for modulated arc
  therapy, 4D RT and,
• Adaptive radiation therapy
• Biological imaging, modeling and biological
  image guided radiation therapy
• Nanotechnology for radiation oncology

91
Physics in Medicine and Biology 50, 23-31, 2005
92
Carbon Nanotube
CNT is a tubular form of carbon with diameter as
small as 1 nm. Length few nm to microns. CNT
is configurationally equivalent to a two
dimensional graphene sheet rolled into a tube.
CNT exhibits extraordinary mechanical properties
Youngs modulus over 1 Tera Pascal, as stiff as
diamond, and tensile strength 200 GPa. CNT can
be metallic or semiconducting, depending on
chirality.
93
(No Transcript)
94
No radiation damage was found in nanotube
structure.
Before Rx
After Rx
95
I-Vg curve before and after 6 cG radiation
96

• Potential Applications
• Radiation Dosimetry.
• Implantable dosimeter ?
• New type of imaging device?
• contrast agents

97
Summary

• 3D 4D CT, CBCT, PET

• 4D RT is emerging.

• Deformable registration, contour mapping,
  accumulated dose calculation, 4D inverse
  planning,
• Biologically conformal radiation therapy.

98
ACKNOWLEDGEMENT

• T. Li, Y. Yang, J. Wang, L. Zhu, M. Chao, R.
  Wiersma, L. Lee, W. Mao, E. Schreibmann, D.
  Paquin, Y. Xie, N. Wink, B. Thorndyke,
• Clinical faculty
• Koong, Q. Le, B. Loo, G Luxton, P. Lee, C. King,
  S. Hancock, P. Keall, P. Maxim

99
Projects

• Comparison of SUVs of 3D and 4D PET

• Small animal functional imaging

• Adaptive therapy replanning
• Biologically conformal radiation therapy.

• Monitoring of tumor motion real-time.