8/2/2012 SRS using the CyberKnife Disclaimer/Conflict of Interest

advertisement
8/2/2012
SRS using the CyberKnife
Sonja Dieterich, PhD, DABR
Associate Professor
University of California Davis
Disclaimer/Conflict of Interest
Consulting agreements with Broncus Medical and
CyberHeart, Inc.
Scientific Advisory Board, MGS Research
4 Essentials of CK Physics
1.
Dose Delivery Accuracy
2.
Small-field Dosimetry for SRS
3.
Patient Localization
4.
Treatment Planning
1
8/2/2012
The CyberKnife System
n
n
n
n
1-30 Fractions in 15-60 min
1000 Mu/min
Image-guidance during
treatment
Corrects:
– 10 mm in translation,
– 1.5 degree roll/pitch angle
– 3 degree yaw angle
n
<0.95 mm accuracy
as defined by E2E test
Image courtesy of Accuray Inc.
1. Dose Delivery Accuracy
Winston-Lutz Test
•Reference:
W. Lutz, K. R. Winston and N. Maleki, "A system for
stereotactic radiosurgery with a linear accelerator,"
Int J Radiat Oncol Biol Phys 14, 373-381 (1988).
Aimed at frame-based
SRS
• Image-based setup not
included
•
2
8/2/2012
Modified Winston-Lutz: AQA
• Modified Winston-Lutz:
• 2 beams
• Image-guided setup using fiducials
• Performed daily
• AQA is lacking:
• Dose distribution is overlay of
many beams
• 4 localization modalities
• Real-time tracking
Modified Winston-Lutz: End-to-End (E2E)
Process & Results of E2E Test
• E2E includes complete
process:
• Simulation
• Contouring
• Planning
• Setup
• Localization
• Treatment delivery
• Specification < 0.95 mm
• Actuals 0.3 mm – 0.7 mm
3
8/2/2012
Delivery Accuracy in Patient
n
n
n
Near real-time image
guidance
Frequency based on
acceptable PTV margin
Small PTV margin vs.
image dose and Tx time
Murphy, M. J. (2009). "Intrafraction Geometric
Uncertainties in Frameless Image-Guided Radiosurgery."
International Journal of Radiation
Oncology*Biology*Physics 73(5): 1364-1368.
Furweger et al 2010
SAMS Question 1
Which test defines the overall accuracy of a
CyberKnife?
20%
1.
20%
2.
20%
3.
20%
4.
20%
5.
Winston-Lutz test
Daily beam pointing accuracy (AQA)
Robot calibration
Delivery QA of patient plan (DQA)
End-to-end test (E2E)
10
C ountdown
SAMS Answer 1
Which test defines the overall accuracy of a CyberKnife?
n
n
n
n
End-to-end (E2E)test
Simulates patient treatment from CT scan through delivery
All localization algorithms in anthropomorphic phantom
0.3 mm – 0.7 mm in clinical practice
References:
Chang, S. D., W. Main, et al. (2003). "An analysis of the accuracy of the CyberKnife: a robotic
frameless stereotactic radiosurgical system." Neurosurgery 52(1): 140-146; discussion 146-147.
Yu, C., W. Main, et al. (2004). "An Anthropomorphic Phantom Study of the Accuracy of CyberKnife
Spinal Radiosurgery." Neurosurgery 55(5): 1138-1149.
4
8/2/2012
2. Small Field Dosimetry
IAEA Formalism for Reference Dosimetry
Alfonso, R., P. Andreo, et al. (2008). "A
new formalism for reference dosimetry of
small and nonstandard fields." Med Phys
35(11): 5179-5186.
TG-51 for Flattening-Filter Free Beam
n
n
TG51 uses %dd(10)x
Details in:
Kalach & Rogers, Med
Phys 30 (2003) 1546
5
8/2/2012
TG 51: How do we get kQ ?
•MC simulation of kQ
• 0.3% difference in linac vs. CK
• Use 6MV linac kQ or measure with 60 mm collimator
•Araki, Med Phys (2006), 2955
TG-51: Measuring %dd(10)x
n CK:
60 mm cylindrical collimator at 80 cm SAD
n Measure
at 10 cm depth, 100 cm SSD, 60 mm collimator
n Calculate
equivalent square
n Interpolate
to 80 cm SAD using the BJR data
TG-51: Chamber selection
Kawachi et al, Med Phys
(2008) 4591
• Dose flatness insufficient for Farmer-type chamber
• Cavity length should be 1 cm or shorter
• Option: cross calibrate a short chamber with Farmer-type chamber
6
8/2/2012
SAMS Question 2
Why does the dosimeter reading of a Farmer-type chamber
need to be corrected for reference dosimetry?
20%
4.
Difference in kQ for flattening filter vs. 6 MV flattening filter
free beam
To adjust for detector alignment uncertainties
Off-axis beam profile for non-flattened beam changes over
chamber length
High dose rate effect (1000 MU/min) on chamber
20%
5.
To adjust for energy-dependent detector response
20%
20%
20%
1.
2.
3.
10
C ountdown
SAMS Answer 2
Why does the dosimeter reading
of a Farmer-type chamber need
to be corrected for reference
dosimetry?
n Off-axis beam profile for nonflattened beam changes over
chamber length
Reference: Kawachi, T., H. Saitoh, et al. (2008).
"Reference dosimetry condition and beam quality
correction factor for CyberKnife beam." Med Phys 35(10):
4591-4598.
3. Patient Localization
7
8/2/2012
The Principle of 2D-3D registration
Fu, D. and G. Kuduvalli (2008). "A fast, accurate, and automatic 2D-3D image registration for image-guided
cranial radiosurgery." Med Phys 35(5): 2180-2194.
Cranial Localization
XSight Spine Localization
D. Fu, G. Kuduvalli, C. J. Maurer, J. Allision and J. Adler, "3D target localization using 2D local displacements of skeletal structures in
orthogonal x-ray images for image-guided spinal radiosurgery," Int J CARS 1, 198-200 (2006).
8
8/2/2012
Spine Localization Accuracy
Fürweger, C., C. Drexler, et
al. (2010). "Patient Motion
and Targeting Accuracy in
Robotic Spinal
Radiosurgery: 260 SingleFraction Fiducial-Free
Cases." International
Journal of Radiation
Oncology*Biology*Physics
78(3): 937-945.
SAMS Question 3
Which tracking algorithm would be used for a
C3 spine target?
20%
1.
20%
2.
20%
3.
20%
4.
20%
5.
Cranial tracking
Spine Tracking
Fiducial tracking
LOT tracking
Spine segmentation
10
C ountdown
SAMS Answer 3
Which tracking algorithm would be used for a C3 spine
target?
n Spine Tracking
n C3 barely visible in DRR
n Highly flexible cervical spine means cranial tracking not
accurate
n Spine segmentation removes mandible interference, allows
for stable tracking
Reference: Fürweger, C., C. Drexler, et al. (2010). "Patient Motion and Targeting Accuracy in
Robotic Spinal Radiosurgery: 260 Single-Fraction Fiducial-Free Cases." International Journal of
Radiation Oncology*Biology*Physics 78(3): 937-945.
9
8/2/2012
4. Treatment Planning
Sequential Optimization Example
1.
Initial optimization step,
optimize the minimum dose in
the PTV
2.
Set a constraint on the minimum
dose, optimize the conformality
3.
Set a constraint on conformality,
optimize total beam weight:
19895 MU → 18567 MU
Sequential Optimization
Schlaefer, A. and A. Schweikard (2008).
"Stepwise multi-criteria optimization for
robotic radiosurgery." Med Phys 35(5):
2094-2103.
10
8/2/2012
Example of Sequential Planning
PDD Calculations for 4 Collimators
with Raytrace and Monte Carlo
• Change in PDD most
pronounced for small
collimators
• SRS near air cavities,
e.g. nasal sinuses
• MC calculation time:
15 - 20 min for 2%
uncertainty at dmax
SAMS Question 4
What is the treatment planning algorithm used to for
inhomogeneity corrections?
20%
1.
20%
2.
20%
3.
20%
4.
20%
5.
Convolution-Superposition
Equivalent tissue-air-ratio
Monte Carlo
Linear Boltzman transport equation
Simulated annealing
10
C ountdown
11
8/2/2012
SAMS Question 4
What is the treatment planning algorithm used to for
inhomogeneity corrections?
n Monte Carlo
n Use might be necessary near air cavities or large metal
implants
n AVMs: CT artefact vs. accuracy of dose calculation?
References:
Wilcox, E. E. and G. M. Daskalov (2008). "Accuracy of dose measurements and calculations
within and beyond heterogeneous tissues for 6 MV photon fields smaller than 4 cm produced by
Cyberknife." Med Phys 35(6): 2259-2266.
Summary
n
n
n
n
CK dose delivery accuracy in E2E meets SRS
standards
Well understood small field reference dosimetry
Integrated image guidance system for frameless
& fiducial-less localization and tracking
MC dose calculation & Sequential optimization
12
Download