Geometrical Splitting Technique to Improve the Computational Efficiency in Monte Carlo Calculations for Proton Therapy
J Ramos-Mendez1*, J Perl2, B Faddegon3, H Paganetti4, (1) Benérita Universidad Autónoma de Puebla, Puebla, México, (2) Stanford Linear Accelerator Center, Menlo Park, CA, (3) UC San Francisco, San Francisco, CA, (4) Massachusetts General Hospital, Boston, MASU-E-T-478 Sunday 3:00:00 PM - 6:00:00 PM Room: Exhibit Hall
Purpose: To implement a geometry based particle splitting technique in order to reduce the computation time when generating treatment head phase space files for proton therapy dose calculations using Monte Carlo (MC) calculations and to validate the doses generated from these phase spaces with respect to reference simulations.
Methods: The treatment nozzles at the Francis H Burr Proton Therapy Center (FHBPTC) were modeled with a new MC tool ('TOPAS' based on Geant4). For variance reduction purposes, two particle-splitting planes were implemented, one downstream of the second ionization chamber the other upstream of the aperture of the nozzle and phase spaces in IAEA format were recovered. The symmetry of the proton beam was considered to split the particles by a factor of 4 per plane. Split particles were randomly positioned at different locations rotated around the beam axis. The computational efficiency was calculated and dose profiles compared for a voxelized water phantom for different treatment fields for both the reference and optimized simulations. Depth-dose curves and beam profiles were analyzed. Dose calculation in patients was simulated to compare the performance.
Results: Normalized computational efficiency between 10 and 14.5 were reached. Percentage difference between dose profiles in water for simulations done with and without particle splitting is within the statistical precision of 2%, 1 standard deviation. Dose distributions for the realistic patient treatment show differences up to 4% in the regions of interest, within 2 standard deviations.
Conclusions: By considering the cylindrically symmetric region of the nozzle and the splitting planes separated at strategic distance, considerable time reduction can be achieved without compromising the precision. This approach will reduce the time for phase space simulations for clinical MC dose calculation at FHBPTC by more than a factor of 10.