基于蒙特卡罗模拟的直肠癌术前容积调强放射治疗计划剂量验证

目的: 通过蒙特卡罗模拟评价基于各向异性解析算法（AAA）直肠癌术前容积调强放射治疗计划（VMAT）的剂量计算精度。方法: 选取20例基于AAA算法和RapidPlan模型优化的直肠癌术前VMAT计划，通过对比蒙卡模拟与治疗计划计算结果的平均DVH、靶区适形度（CI）、靶区均匀性（HI）和Gamma 3D通过率等参数，评估基于AAA算法的VMAT治疗计划剂量计算精度。结果: 两种Gamma 3D评估策略通过率的均值与标准差分别为97.58±0.47%（Max Dose）、92.46±1.76%（Local Dose），且差异具有统计学意义（P<0.05）；对PTV和PGTV的CI、DMin、膀胱的D50%、DMean等不符合正态分布的参数做相关样本非参数检验，除PGTV的CI和Dmin外，差异均具有统计学意义（P < 0.05）；其他服从正态分布的参数做配对样本T检验，差异均具有统计学意义（P < 0.05）。结论: Rapidplan模型计划在低剂量区通过率较低，说明AAA算法射野边缘低剂量区计算偏差较大；靶区Dmin与算法的精度较为相关，CI和HI参数相对于蒙卡模拟结果有一定差异；AAA算法在股骨头和膀胱的D50%、DMean相对蒙卡模型有不同程度的低估。

Dosimetry Verification of Volume Modulated Radiotherapy Plans for Preoperative Rectal Cancer based on Monte Carlo Simulation

Purpose This work aims to carry out a Monte Carlo based investigation on the precision of dose estimation for Volumetric Modulated Arc Therapy (VMAT) plans based on anisotropic analytical algorithm, focusing on preoperative rectum tumor. Methods 20 VMAT plans of preoperative rectum tumor have been optimized using AAA and RapidPlan model. The precision of RapidPlan estimated dose estimations are evaluated via comparison with independent calculation of dose distribution based on Monte Carlo simulations. The comparison is carried out using several key parameters: mean DVH, Homogeneity Index (HI), Conformity Index (CI), 3D Gamma pass rate. Results 3D Gamma pass rate under two different criteria are 97.58±0.47%、92.46±1.76% for max dose and local dose, respectively，where the difference is statistically significant (P < 0.05) for parameters that do not conform with normal distribution (except for CI of PGTV and Dmin of PTV, where non-parametric test has shown statistical significant difference (P < 0.05) other parameters tested with pair sample t-test all reveal statistical significant difference (P < 0.05).Conclusion The RapidPlan estimation mainly differs in low dose region with MC simulation, which is related to the anisotropic analytical algorithm precision. Mean values of CI and HI in MC simulation are higher than those in RapidPlan estimation. In the meantime, RapidPlan model underestimates Dmean and D50% of femoral head and bladder