Center for Molecular Imaging and Experimental Radiotherapy, Institut de Recherche Expérimentale et Clinique, Université catholique de Louvain, 1200, Brussels, Belgium.
Ion Beam Applications s.a., 1348, Louvain-la-Neuve, Belgium.
Med Phys. 2019 Oct;46(10):4676-4684. doi: 10.1002/mp.13749. Epub 2019 Aug 31.
Proton therapy is very sensitive to treatment uncertainties. These uncertainties can induce proton range variations and may lead to severe dose distortions. However, most commercial tools only offer a limited integration of these uncertainties during treatment planning. In order to verify the robustness of a treatment plan, this study aims at developing a comprehensive Monte Carlo simulation of the treatment delivery, including the simulation of setup and range errors, variation of the breathing motion, and interplay effect.
Most clinically relevant uncertainties have been modeled and implemented in the fast Monte Carlo dose engine MCsquare. Especially, variation of the breathing motion is taken into account by deforming the initial Four-dimensional computed tomography (4DCT) series and generating multiple new 4DCT series with scaled motion. Systematic and random errors are randomly sampled, following a Monte Carlo approach, to generate individual erroneous treatment scenarios. The robustness of treatment plans is analyzed and reported with dose-volume histogram (DVH) bands. The statistical uncertainty coming from the Monte Carlo scenario sampling is studied.
A validation demonstrated the ability of the motion model to generate new 4DCT series with scaled motion amplitude and improved image quality in comparison to the initial 4DCT. The robustness analysis is applied to a lung tumor treatment. Considering the proposed uncertainty model, the simulation of 300 treatment scenarios was necessary to reach an acceptable level of statistical uncertainty on the DVH band.
A comprehensive and statistically sound method of treatment plan robustness verification is proposed. The uncertainty model presented in this paper is not specific to protons and can also be applied to photon treatments. Moreover, the generated 4DCT series, with scaled motion, can be imported in commercial TPSs.
质子治疗对治疗不确定性非常敏感。这些不确定性会导致质子射程变化，并可能导致严重的剂量失真。然而，大多数商业工具在治疗计划中仅提供这些不确定性的有限整合。为了验证治疗计划的稳健性，本研究旨在开发一种全面的质子治疗交付蒙特卡罗模拟，包括设置和射程误差、呼吸运动变化和相互作用效应的模拟。
已对最相关的临床不确定性进行建模并在快速蒙特卡罗剂量引擎 MCsquare 中实现。特别是，通过变形初始四维计算机断层扫描（4DCT）系列并生成具有缩放运动的多个新 4DCT 系列来考虑呼吸运动的变化。系统和随机误差根据蒙特卡罗方法进行随机采样，以生成单个错误的治疗方案。使用剂量体积直方图（DVH）带分析和报告治疗计划的稳健性。研究了来自蒙特卡罗方案抽样的统计不确定性。
验证表明，运动模型能够生成具有缩放运动幅度的新 4DCT 系列，并与初始 4DCT 相比提高了图像质量。稳健性分析应用于肺肿瘤治疗。考虑到所提出的不确定性模型，需要模拟 300 个治疗方案，才能在 DVH 带的统计不确定性方面达到可接受的水平。
提出了一种全面且具有统计意义的治疗计划稳健性验证方法。本文提出的不确定性模型不仅适用于质子，也可应用于光子治疗。此外，生成的具有缩放运动的 4DCT 系列可导入商业 TPS。
