| Summary: | 碩士 === 國立清華大學 === 核子工程與科學研究所 === 107 === The understanding of proton beams’ properties and the development of corresponding source models are the critical initial point of Monte Carlo (MC) application for proton therapy. The purpose of this study is to discuss the Monte Carlo (MC) source models developed for pencil beam scanning (PBS) proton machine. By series of comparison between simulation and measurement, this study has developed the best source model among the published literature. The construction of the source parameters, testing and verified results will be described in detail in this article.
Among the beam delivery techniques in proton therapy. Unlike wobbling and passive scattering techniques, proton beams in PBS do not require scattering process to obtain an uniform dose distribution in tumor; therefore, the simulation of the source term is not severely susceptible to components inside the beam nozzle. Yet since the uniform dose distribution delivered by PBS consists of numerous proton beams each with unique property, accurate description of each beam’s property is a necessity. This study first compares the dose distribution calculated for proton therapy using various MC codes including FLUKA、GEANT4、MCNP and PHITS. Ensuring that, in the case of a standard 10×10 cm^2 field, the gamma passing rates (3%/3mm) of all four codes are all above 95%. Eventually, GEANT4 is used in the following study. The development of the source models in this study is based on the monthly QA results from proton therapy center.
This study has proposed a souce model that could almost perfectly reproduce the QA results. We then compared this source model with other two models in the literature in-depth. The three source models show extremely obvious discrepancies in the case of single beam simulation; yet in the reproduction QA results, the proposed model is able to present the best agreement with the experimental data. Although in simple treatment cases, all three models have over 92% gamma passing rate (3%/3mm); finally in the case considering collimators, the proposed model shows the best agreement with the commercial MC-based Treatment Planning System (TPS) and experimental data. For the development of proton therapy technology in the future, for instance, extremely small-size proton beam or cases with complex collimators, we could expected that the effectiveness of this souce model would be emphasized.
|
|---|
