Biomedical Engineering, University of Calgary, Calgary, Alberta, Canada.
Tom Baker Cancer Centre, Calgary, Alberta, Canada.
Med Phys. 2022 Jun;49(6):3585-3596. doi: 10.1002/mp.15651. Epub 2022 Apr 28.
The purpose of this analysis is to predict worsening post-treatment normal tissue toxicity in patients undergoing accelerated partial breast irradiation (APBI) therapy and to quantitatively identify which diagnostic, anatomical, and dosimetric features are contributing to these outcomes.
A retrospective study of APBI treatments was performed using 32 features pertaining to various stages of the patient's treatment journey. These features were used to inform and construct a Bayesian network (BN) based on both statistical analysis of feature distributions and relative clinical importance. The target feature for prediction was defined as a measurable worsening of telangiectasia, subcutaneous tissue induration, or fibrosis when compared against the observed baseline. Parameter learning for the network was performed using data from the 299 patients included in the ACCEL trial and predictive performance was measured. Feature importance for the BN was quantified using a novel information-theoretic approach.
Cross-validated performance of the BN for predicting toxicity was consistently higher when compared against conventional machine learning (ML) techniques. The measured BN receiver operating characteristic area under the curve was 0.960 0.013 against the best ML result of 0.942 0.021 using five-fold cross-validation with separate test data across 100 trials. The volume of the clinical target volume, gross target volume, and baseline toxicity measurements were found to have the highest feature importance and mutual dependence with normal tissue toxicity in the network, representing the strongest contribution to patient outcomes.
The BN outperformed conventional ML techniques in predicting tissue toxicity outcomes and provided deeper insight into which features are contributing to these outcomes.
本分析旨在预测行加速部分乳腺照射（APBI）治疗的患者治疗后正常组织毒性恶化，并定量确定哪些诊断、解剖和剂量学特征对此类结果有影响。
对 APBI 治疗进行了回顾性研究，使用了与患者治疗过程各个阶段相关的 32 个特征。这些特征用于根据特征分布的统计分析和相对临床重要性来构建和构建基于贝叶斯网络（BN）。预测的目标特征定义为与观察到的基线相比，毛细血管扩张、皮下组织硬结或纤维化的可测量恶化。使用 ACCEL 试验中包含的 299 名患者的数据进行网络参数学习，并测量预测性能。使用一种新的信息论方法对 BN 的特征重要性进行量化。
与传统机器学习（ML）技术相比，BN 预测毒性的交叉验证性能始终更高。使用 100 次试验的 5 倍交叉验证和单独的测试数据，BN 的测量接收器操作特征曲线的 AUC 为 0.960 0.013，而最佳 ML 结果为 0.942 0.021。临床靶区体积、大体靶区体积和基线毒性测量值在网络中与正常组织毒性的特征重要性和相互依赖性最高，代表对患者结果的最大贡献。
BN 在预测组织毒性结果方面优于传统的 ML 技术，并深入了解哪些特征对这些结果有影响。
