2023 Volume 12 Issue 1 Pages 53-64
We validated the calculation accuracies and times between the pencil beam algorithm (PBA) and the Monte Carlo (MC) algorithm in the new proton pencil beam scanning system. Thirty-three (homogeneous phantom) and three (heterogeneous phantom) uniform-dose plans were verified for cubic targets. These plans were calculated using the PBA and five MC statistical uncertainties of 0.3%, 0.5%, 1.0%, 1.5%, and 2.0%. We also evaluated the required dose calculation times per beam with the statistical analysis. Then, eight clinically realistic beams were validated as the end-toend test. The dose differences of 0.3% and 0.5% uncertainty MC plans were satisfied our tolerance (< ±3.0%). All five uncertainty MC plans improved at least the average/minimum gamma score from 93.8%/33.5% to 98.5%/88.9% for homogeneous (2%/2 mm) and 96.7%/83.9% to 98.8%/94.1% for heterogeneous (3%/3 mm) than PBA plans. However, the average calculation times of 0.3% and 0.5% uncertainties in MC plans were 15.5 and 5.8 times longer than in PBA plans (p < 0.001). The endto-end tests satisfied the acceptable with a statistical uncertainty of MC below 0.5%. Although the dose calculation times would be significantly increased, we suggest that the statistical uncertainty of MC below 0.5% is appropriate for clinical use with PBS plans.