How Do We Treat a Moving Tumor?
Fluoroscopic Lung Tumor
Tracking without Implanted
Fiducial Markers
Steve B. Jiang, Ph.D.
Depart Of Radiation Oncology
University of California San Diego
sbjiang@ucsd.edu
radonc.ucsd.edu/Research/CART
2
Tracking based on External Surrogates
Shooting A Target by Looking in The Mirror
Image in the mirror
 Target
External surrogates
 Tumor
Varian RPM
CyberKnife Synchrony
3
4
Tracking using Implanted Fiducial Markers
Gating based on External Surrogates
Image Intensifier
100
90
Target Coverage (%)
80
70
60
50
40
30
20
10
0
1
4
5
6
7
8
9
10
11
avg
Patients
Kanoulas, Aslam, Sharp, Berbeco, Nishioka, Shirato, and Jiang
Phys Med Biol.52(17):5443-55456, 2007
Wu, Zhao, Berbeco, Nishioka, Shirato, Jiang
Phys Med Biol. 53(24):7137-7150, 2008
X-ray Tube
Calypso 4D Localization System
Mitsubishi/Hokkaido RTRT System
5
6
1
Direct Lung Tumor Tracking w/o Markers
Implantation of Fiducial Markers
 Multiple Template Tracking
18G x 150 mm
depth
mark
depth mark

baseof
of needle
needle
base
mandril
Cui, Dy, Sharp, Alexander, and
Jiang, Phys Med Biol, 52(20):62296242, 2007.
 Active Shape Model Tracking
CT guided implantation
0.8
mmφmmφ
0.6～0.8
31.5～3
mm mm

Xu, Hamilton, Schowengerdt, and
Jiang, Phys Med Biol, 52(17):52775293, 2007.
 Optical Flow Tracking

US guided implantation
7
8
Indirect Fluoroscopic Tracking


Xu, Hamilton, Schowengerdt,
Alexander, and Jiang, Med Phys,
35(12):5351-9, 2008.
Indirect Fluoroscopic Tracking
Issues with direct tracking

Directly tracking of the tumor
sometimes is impossible due to
poor image quality and low
target contrast

Common tracking methods
used in computer vision often
fails since tumor has no color,
no texture, and often no clear
shape
We proposed an indirect
tracking approach

Tracking invisible tumor by
tracking visible surrogate
features
Lin, Cervino, Tang, Vasconcelos, and Jiang, Phys Med Biol 54(4):
981-992, 2009.
9
10
Diaphragm as an Internal Surrogate
Diaphragm Does Not Always Work
6
True position
Model 1
Model 2
4
Displacement (mm)
2
0
-2
-4
-6
-8
0
5
10
15
20
25
30
Time (s)
Mean error ~ 1 mm
Maximum error (e95) ~ 2 mm
Cervino, Chao, Sandhu, and Jiang
Phys Med Biol 54(11): 3529-3541, 2009
11
12
2
Future work: building a comprehensive tracking system
Tumor Tracking during Arc Therapy
Surface Tracking Camera
X-ray
(Always on)
(Only when necessary)
Courtesy of Elekta
13
14
Reconstruct Tumor Trajectory from CBCT Projections
Tumor Tracking in Rotational Geometry

Reference templates

Volume of Interest (VOI)



DRRs generated for each phase of 4DCT
VOI generated for 4DCT
VOI projected, defining an ROI in each DRR and CBCT projection
Reference Templates
15
16
Phase Binned Trajectory Reconstruction

CBCT Projections
Preliminary Results: Digitial Phantom (NCAT)
Draw lines from each source position to each tumor
position
True Position
Algorithm Output
Tumor Position (mm)
Phase bin each line based on
breathing signal
AP (cm)
LR
(
cm
)
AP (cm)
SI (cm)

LR (cm)
17
Projection Angle
18
3
Preliminary Results: Physical Phantom
Acknowledgement
 John Lewis
 Ruijiang Li
 Laura Cervino
 Ajay Sandhu
19
20
4