Program Information
Individual Beam QA for a Robotic Radiosurgery System Using a Scintillator Cone
C McGuinness*, M Descovich , A Sudhyadhom , University of California San Francisco, San Francisco, CA
Presentations
SU-G-TeP4-12 (Sunday, July 31, 2016) 5:30 PM - 6:00 PM Room: ePoster Theater
Purpose:
The targeting accuracy of the Cyberknife system is measured by end-to-end tests delivering multiple isocentric beams to a point in space. While the targeting accuracy of two representative beams can be determined by a Winston-Lutz-type test, no test is available today to determine the targeting accuracy of each clinical beam. We used a scintillator cone to measure the accuracy of each individual beam.
Methods:
The XRV-124 from Logos Systems Int'l is a scintillator cone with an imaging system that is able to measure individual beam vectors and a resulting error between planned and measured beam coordinates. We measured the targeting accuracy of isocentric and non-isocentric beams for a number of test cases using the Iris and the fixed collimator.
Results:
Collimator---------------Num. Beams-------Avg. Diff. (mm)------StDev. (mm)--------Max Diff. (mm)
Fixed 10mm____________52___________0.837____________0.380___________1.73
Iris 10mm______________52___________1.076____________0.433___________1.873
Iris 12.5mm____________52___________1.232____________0.489___________2.043
Iris 15mm______________52___________1.218____________0.475___________2.01
Iris 20mm______________52___________1.202____________0.501___________2.508
Iris 25mm______________52___________1.234____________0.494___________2.32
Iris 30mm______________52___________1.096____________0.483___________2.433
PlanQA Fixed 20mm______86___________0.900____________0.388___________1.667
PlanQA Iris 12.5-30mm___164__________0.806____________0.313___________1.571
The average difference between plan and measured beam position was 0.8-1.2mm across the collimator sizes and plans considered here. The max error for a single beam was 2.5mm for the isocentric plans, and 1.67mm for the non-isocentric plans. The standard deviation of the differences was 0.5mm or less.
Conclusion:
The CyberKnife System is specified to have an overall targeting accuracy for static targets of less than 0.95mm. In E2E tests using the XRV124 system we measure average beam accuracy between 0.8 to 1.23mm, with maximum of 2.5mm. We plan to investigate correlations between beam position error and robot position, and to quantify the effect of beam position errors on patient specific plans.
Funding Support, Disclosures, and Conflict of Interest: Martina Descovich has received research support and speaker honoraria from Accuray
Contact Email: