Interpretable machine learning for predicting pathologic complete response in patients treated with chemoradiation therapy for rectal adenocarcinoma.
Interpretable machine learning for predicting pathologic complete response in patients treated with chemoradiation therapy for rectal adenocarcinoma.
复制标题
DOI：
10.3389/frai.2022.1059033
复制
发表时间：
2022
影响因子：
4
通讯作者：
Xiao, Ying
中科院分区：
文献类型：
--
作者：
Wang, Du;Lee, Sang Ho;Geng, Huaizhi;Zhong, Haoyu;Plastaras, John;Wojcieszynski, Andrzej;Caruana, Richard;Xiao, Ying
关键词：
病理完全缓解（pCR）是决定直肠癌（RC）患者在新辅助放化疗（nCRT）后是否应该进行手术的关键因素。目前，病理学家对手术标本的组织学分析是可靠评估pCR的必要条件。机器学习（ML）算法有可能成为一种非侵入性的方法，用于识别非手术治疗的合适候选人。然而，这些机器学习模型的可解释性仍然具有挑战性。我们建议使用可解释增强机（EBM）预测nCRT后RC患者的pCR。共提取296个特征，包括临床参数（CPs）、总肿瘤体积（GTV）和危险器官的剂量-体积直方图（DVH）参数，以及GTV的放射组学(R)和剂量组学(D)特征。R和D特征被细分为形状(S)、一阶（L1）、二阶（L2）和高阶（L3）局部纹理特征。采用多视图分析确定最佳输入特征类别集。Boruta用于为每个输入数据集选择所有相关特征。ML模型在我院180例病例上进行训练，其中37例RTOG 0822临床试验作为模型验证的独立数据集。使用ROC AUC评估EBM在测试数据集上预测pCR的性能，并与三种最先进的黑箱模型（极端梯度增强（XGB），随机森林（RF）和支持向量机（SVM））进行比较。所有黑箱模型的预测都使用沙普利加性解释进行解释。所有模型的最佳输入特征类别为CP+DVH+S+R_L1+R_L2，其中boruta选择的特征使EBM、XGB、RF和SVM模型的auc分别达到0.820、0.828、0.828和0.774。虽然EBM没有达到最佳效果，但它提供了识别不同特征值反应评分关键转折点的最佳能力，揭示了最大剂量>50 Gy的膀胱，以及最大直径>80 mm，延伸率<0.55，最小轴长>50 mm和CT强度方差较小的肿瘤与不良预后相关。EBM有可能提高医生评估基于ml的pCR预测的能力，并对选择患者采取“观察等待”策略进行RC治疗具有重要意义。
Pathologic complete response (pCR) is a critical factor in determining whether patients with rectal cancer (RC) should have surgery after neoadjuvant chemoradiotherapy (nCRT). Currently, a pathologist's histological analysis of surgical specimens is necessary for a reliable assessment of pCR. Machine learning (ML) algorithms have the potential to be a non-invasive way for identifying appropriate candidates for non-operative therapy. However, these ML models' interpretability remains challenging. We propose using explainable boosting machine (EBM) to predict the pCR of RC patients following nCRT. A total of 296 features were extracted, including clinical parameters (CPs), dose-volume histogram (DVH) parameters from gross tumor volume (GTV) and organs-at-risk, and radiomics (R) and dosiomics (D) features from GTV. R and D features were subcategorized into shape (S), first-order (L1), second-order (L2), and higher-order (L3) local texture features. Multi-view analysis was employed to determine the best set of input feature categories. Boruta was used to select all-relevant features for each input dataset. ML models were trained on 180 cases from our institution, with 37 cases from RTOG 0822 clinical trial serving as the independent dataset for model validation. The performance of EBM in predicting pCR on the test dataset was evaluated using ROC AUC and compared with that of three state-of-the-art black-box models: extreme gradient boosting (XGB), random forest (RF) and support vector machine (SVM). The predictions of all black-box models were interpreted using Shapley additive explanations. The best input feature categories were CP+DVH+S+R_L1+R_L2 for all models, from which Boruta-selected features enabled the EBM, XGB, RF, and SVM models to attain the AUCs of 0.820, 0.828, 0.828, and 0.774, respectively. Although EBM did not achieve the best performance, it provided the best capability for identifying critical turning points in response scores at distinct feature values, revealing that the bladder with maximum dose >50 Gy, and the tumor with maximum2DDiameterColumn >80 mm, elongation <0.55, leastAxisLength >50 mm and lower variance of CT intensities were associated with unfavorable outcomes. EBM has the potential to enhance the physician's ability to evaluate an ML-based prediction of pCR and has implications for selecting patients for a “watchful waiting” strategy to RC therapy.
登录
查看更多内容
影响因子：
3.5
作者：
Lee, Sang Ho;Han, Peijin;Lee, Junghoon
通讯作者：
Lee, Junghoon
影响因子：
8.9
作者：
Crimi, Filippo;Capelli, Giulia;Stramare, Roberto
通讯作者：
Stramare, Roberto
影响因子：
5.8
作者：
Kursa, Miron B.;Rudnicki, Witold R.
通讯作者：
Rudnicki, Witold R.
基于深度学习的放射组学模型，用于预测多形性胶质母细胞瘤的生存期
影响因子：
4.6
作者：
Lao J;Chen Y;Li ZC;Li Q;Zhang J;Liu J;Zhai G
通讯作者：
Zhai G
影响因子：
4.8
作者：
Chen, Fangying;Ma, Xiaolu;Lu, Jianping
通讯作者：
Lu, Jianping