TrueBeam直线加速器蒙特卡罗模拟的几何模型
A geometrical model for the Monte Carlo simulation of the TrueBeam linac
摘要：
直线加速器的蒙特卡罗（MC）模拟依赖于加速器头部的精确几何描述。Varian TrueBeam（TB）直线加速器的几何结构对研究人员来说是不可用的。相反，该公司分发相空间文件（PSF）的无滤波器（FFF）光束计数上游的钳口。然而，基于第三方统计的PSF的MC模拟受到限制。我们提出了一个实验为基础的几何开发的模拟的TB直线加速器的FFF光束。TB直线加速器的上部被建模，修改了Clinac 2100的几何形状。最重要的修改是由特设薄过滤器，通过比较剂量测量和模拟建模的标准平整过滤器的替代。6 MV和10 MV FFF射束的实验剂量分布是从Varian Golden数据集和辐射场范围为3X 3至40 X40 cm 2的内部测量中获得的。与建议的几何模型得到的实验数据和模拟结果之间的协议指标的剂量差异，均方根误差和伽马指数。对使用Varian为TB直线加速器分配的第二代PSF从MC模拟获得的剂量分布进行了相同的比较。比较结果显示，对于所考虑的所有射野，模拟几何形状的剂量与使用TB PSF模拟获得的剂量相似，但40 X40 cm 2射野除外，其中模拟几何形状能够更准确地再现测量剂量。我们的方法使得有可能：（i）调整初始束参数，以匹配测量的剂量分布;（ii）减少统计不确定性到任意低的值;和（iii）评估系统的不确定性，采用不同的MC代码。
英文摘要：
Monte Carlo (MC) simulation of linacs depends on the accurate geometrical description of the head. The geometry of the Varian TrueBeam (TB) linac is not available to researchers. Instead, the company distributes phase-space files (PSFs) of the flattening-filter-free (FFF) beams tallied upstream the jaws. Yet, MC simulations based on third party tallied PSFs are subject to limitations. We present an experimentally-based geometry developed for the simulation of the FFF beams of the TB linac. The upper part of the TB linac was modeled modifying the Clinac 2100 geometry. The most important modification is the replacement of the standard flattening filters by ad hoc thin filters which were modeled by comparing dose measurements and simulations. The experimental dose profiles for the 6MV and 10MV FFF beams were obtained from the Varian Golden Data Set and from in-house measurements for radiation fields ranging from 3X3 to 40X40 cm2. Indicators of agreement between the experimental data and the simulation results obtained with the proposed geometrical model were the dose differences, the root-mean-square error and the gamma index. The same comparisons were done for dose profiles obtained from MC simulations using the second generation of PSFs distributed by Varian for the TB linac. Results of comparisons show a good agreement of the dose for the ansatz geometry similar to that obtained for the simulations with the TB PSFs for all fields considered, except for the 40X40 cm2 field where the ansatz geometry was able to reproduce the measured dose more accurately. Our approach makes possible to: (i) adapt the initial beam parameters to match measured dose profiles; (ii) reduce the statistical uncertainty to arbitrarily low values; and (iii) assess systematic uncertainties by employing different MC codes.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
