Developing a Workflow Using Supercomputer Accelerated Monte Carlo Simulation Based On Machine Log-Files for Patient Specific Quality Assurance of the Mevion S250i Proton Therapy System
B Chiang*, Y Chen, S Ahmad, University of Oklahoma Health Sciences Center, Oklahoma City, OK
Presentations
PO-GePV-T-72 (Sunday, 7/25/2021) [Eastern Time (GMT-4)]
Purpose: The measurement based proton specific quality assurance (PSQA), a time-consuming process, plays an important role for proton therapy. The purpose of this work has been to develop a unique workflow of running machine log-files based Monte Carlo simulation on supercomputer for PSQA of the Mevion S250i proton therapy system. The proposed method not only takes the delivery errors into account but also saves personnel and machine beam-on time.
Methods: The Mevion beamline was modeled in detail using TOPAS Monte Carlo code. The machine log file was acquired during delivery and converted by our in-house batch file converter to create the input file for running machine log-files based Monte Carlo simulation on supercomputer. The 3D dose was then reconstructed in a solid water phantom. Similar to measurement based PSQA, the 2D Gamma analysis was conducted in three different depths that include isocenter, proximal and distal fall-off and compared with the corresponding doses calculated from the treatment planning system.
Results: Simulation time for 10^8 protons on 5×5×5 cm^3 uniform cubic plan was reduced from almost 2 days using 10 cores on the local server to 1.9 hours using 20 compute nodes on supercomputer. The 2D dose profile comparison at three different depths showed good agreement between simulation and treatment planning system. The Gamma index passing rates for both one beam 5×5×5 cm^3 uniform cubic plan and two beam IROC pelvis plan were passed more than 95% satisfying 4%/4mm criteria.
Conclusion: A unique workflow of how to utilize supercomputer for log-file based Monte Carlo simulation has been developed, which can improve both efficiency and authenticity of utilizing Monte Carlo simulation for PSQA of proton therapy within the clinically acceptable level.
ePosters
Keywords
Protons, Monte Carlo, Quality Assurance
Taxonomy
TH- External Beam- Particle/high LET therapy: Proton therapy – quality assurance
Contact Email