Beam data modeling of linear accelerators (linacs) through machine learning and its potential applications in fast and robust linac commissioning and quality assurance.
Beam data modeling of linear accelerators (linacs) through machine learning and its potential applications in fast and robust linac commissioning and quality assurance.
复制标题
DOI：
10.1016/j.radonc.2020.09.057
复制
发表时间：
2020-12
期刊：
影响因子：
--
通讯作者：
Schüler E
中科院分区：
文献类型：
--
作者：
Zhao W;Patil I;Han B;Yang Y;Xing L;Schüler E
提出一种基于机器学习的新颖方法，用于对直线加速器光束数据进行可靠且准确的建模，适用于直线加速器调试和质量保证过程。我们假设射束数据是固有直线加速器特征和百分比深度剂量 (PDD) 的函数，并且不同射野大小的轮廓彼此相关。使用机器学习框架将相关性表述为多变量回归问题。使用从多个机构获取的 Varian TrueBeam 光束数据集 (n=43) 来评估该框架。数据集包括不同能量和场大小的 PDD 和剖面。使用 10×10cm2 视场作为输入，训练多元回归模型来预测光束特定的 PDD 和不同视场大小的轮廓。在所研究的不同光束能量中，PDD 的预测平均绝对百分比相对误差 (%RE) 为 0.19-0.35%。最大平均绝对 %RE 为 0.93%。对于轮廓预测，平均绝对 %RE 为 0.66–0.93%，最大绝对 %RE 为 3.76%。 PDD 和剖面预测的最大不确定性分别出现在构建区域和现场半影处。随着训练集数量增加到 20 个左右，预测精度不断提高。通过这种基于机器学习的新颖方法，我们展示了用于常规放射治疗直线加速器调试的射束数据的准确且可重复的生成。该方法有可能简化直线加速器调试程序，节省时间和人力，同时提高调试过程的准确性。
To propose a novel machine learning-based method for reliable and accurate modeling of linac beam data applicable to the processes of linac commissioning and QA. We hypothesize that the beam data is a function of inherent linac features and percentage depth doses (PDDs) and profiles of different field sizes are correlated with each other. The correlation is formulated as a multivariable regression problem using a machine learning framework. Varian TrueBeam beam data sets (n=43) acquired from multiple institutions were used to evaluate the framework. The data sets included PDDs and profiles across different energies and field sizes. A multivariate regression model was trained for prediction of beam specific PDDs and profiles of different field sizes using a 10×10cm2 field as input. Predictions of PDDs were achieved with a mean absolute percent relative error (%RE) of 0.19–0.35% across the different beam energies investigated. The maximum mean absolute %RE was 0.93%. For profile prediction, the mean absolute %RE was 0.66–0.93% with a maximum absolute %RE of 3.76%. The largest uncertainties in the PDD and profile predictions were found at the build-up region and at the field penumbra, respectively. The prediction accuracy increased with the number of training sets up to around 20 training sets. Through this novel machine learning-based method we have shown accurate and reproducible generation of beam data for linac commissioning for routine radiation therapy. This method has the potential to simplify the linac commissioning procedure, save time and manpower while increasing the accuracy of the commissioning process.
登录
查看更多内容
影响因子：
3.8
作者：
Almond, PR;Biggs, PJ;Rogers, DWO
通讯作者：
Rogers, DWO
影响因子：
0.4
作者：
Al Mashud, Abdullah;Tariquzzaman, M.;Zakaria, Ga
通讯作者：
Zakaria, Ga
DOI：
10.1016/j.ijrobp.2007.08.083
发表时间：
2008-01-01
影响因子：
7
作者：
Ibbott, Geoffrey S.;Followill, David S.;Roll, Joys E.
通讯作者：
Roll, Joys E.
影响因子：
3.8
作者：
KOSUNEN, A;ROGERS, DWO
通讯作者：
ROGERS, DWO
影响因子：
3.8
作者：
Liu, Paul Zhi Yuan;O'Brien, Ricky;Keall, Paul
通讯作者：
Keall, Paul