Program Information
Validation of a Commercial Monte Carlo Dose Calculation Algorithm for Proton Therapy
J Saini1*, S St. James2 , E Traneus3 , T Wong1 , R Stewart2 , C Bloch2 , (1) SCCA Proton Therapy Center, Seattle, WA, (2) University of Washington, Seattle, WA, (3) RAYSEARCH LABORATORIES AB, Stockholm.
Presentations
SU-F-T-155 (Sunday, July 31, 2016) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose:Compare proton pencil beam scanning dose measurements to GATE/GEANT4 (GMC) and RayStation™ Monte Carlo (RMC) simulations.
Methods:Proton pencil beam models of the IBA gantry at the Seattle Proton Therapy Center were developed in the GMC code system and a research build of the RMC. For RMC, a preliminary beam model that does not account for upstream halo was used. Depth dose and lateral profiles are compared for the RMC, GMC and a RayStation™ pencil beam dose (RPB) model for three spread out Bragg peaks (SOBPs) in homogenous water phantom. SOBP comparisons were also made among the three models for a phantom with a (i) 2 cm bone and a (ii) 0.5 cm titanium insert.
Results:Measurements and GMC estimates of R80 range agree to within 1 mm, and the mean point-to-point dose difference is within 1.2% for all integrated depth dose (IDD) profiles. The dose differences at the peak are 1 to 2%. All of the simulated spot sigmas are within 0.15 mm of the measured values. For the three SOBPs considered, the maximum R80 deviation from measurement for GMC was -0.35 mm, RMC 0.5 mm, and RPB -0.1 mm. The minimum gamma pass using the 3%/3mm criterion for all the profiles was 94%. The dose comparison for heterogeneous inserts in low dose gradient regions showed dose differences greater than 10% at the distal edge of interface between RPB and GMC. The RMC showed improvement and agreed with GMC to within 7%.
Conclusion:The RPB dosimetry show clinically significant differences (> 10%) from GMC and RMC estimates. The RMC algorithm is superior to the RPB dosimetry in heterogeneous media. We suspect modelling of the beam’s halo may be responsible for a portion of the remaining discrepancy and that RayStation will reduce this discrepancy as they finalize the release.
Funding Support, Disclosures, and Conflict of Interest: Erik Traneus is employed as a Research Scientist at RaySearch Laboratories. The research build of the RayStation TPS used in the study was made available to the SCCA free of charge. RaySearch did not provide any monetary support other than a license to use the research build of the TPS.
Contact Email: