Department of Physiology, Pharmacology and Medical Physics, Faculty of Medicine, Alborz University of Medical Sciences, Karaj, Iran.
Department of Radiology, Cancer Institute of Iran, Tehran University of Medical Sciences, Tehran, Iran.
Phys Med. 2019 Jun;62:111-119. doi: 10.1016/j.ejmp.2019.03.013. Epub 2019 May 15.
The aim of this study was to investigate and validate the performance of individual and ensemble machine learning models (EMLMs) based on magnetic resonance imaging (MRI) to predict neo-adjuvant chemoradiation therapy (nCRT) response in rectal cancer patients. We also aimed to study the effect of Laplacian of Gaussian (LOG) filter on EMLMs predictive performance.
98 rectal cancer patients were divided into a training (n = 53) and a validation set (n = 45). All patients underwent MRI a week before nCRT. Several features from intensity, shape and texture feature sets were extracted from MR images. SVM, Bayesian network, neural network and KNN classifiers were used individually and together for response prediction. Predictive performance was evaluated using the area under the receiver operator characteristic (ROC) curve (AUC).
Patients' nCRT responses included 17 patients with Grade 0, 28 with Grade 1, 34 with Grade 2, and 19 with Grade 3 according to AJCC/CAP pathologic grading. In without preprocessing MR Image the best result was for Bayesian network classifier with AUC and accuracy of 75.2% and 80.9% respectively, which was confirmed in the validation set with an AUC and accuracy of 74% and 79% respectively. In EMLMs the best result was for 4 (SVM.NN.BN.KNN) classifier EMLM with AUC and accuracy of 97.8% and 92.8% in testing and 95% and 90% in validation set respectively.
In conclusion, we observed that machine learning methods can used to predict nCRT response in patients with rectal cancer. Preprocessing LOG filters and EL models can improve the prediction process.
本研究旨在探讨和验证基于磁共振成像（MRI）的个体机器学习模型（SVM、贝叶斯网络、神经网络和 KNN 分类器）和集成机器学习模型（ELM）预测直肠癌患者新辅助放化疗（nCRT）反应的性能。我们还旨在研究拉普拉斯高斯（LOG）滤波器对 ELM 预测性能的影响。
98 例直肠癌患者分为训练集（n=53）和验证集（n=45）。所有患者在 nCRT 前一周接受 MRI 检查。从强度、形状和纹理特征集中提取了多个特征。SVM、贝叶斯网络、神经网络和 KNN 分类器分别和联合用于反应预测。使用接收器工作特征（ROC）曲线下面积（AUC）评估预测性能。
根据 AJCC/CAP 病理分级，患者的 nCRT 反应包括 17 例 0 级、28 例 1 级、34 例 2 级和 19 例 3 级。在未经预处理的 MRI 图像中，贝叶斯网络分类器的 AUC 和准确率分别为 75.2%和 80.9%，在验证集的 AUC 和准确率分别为 74%和 79%，结果最佳。在 ELM 中，4（SVM、NN、BN、KNN）分类器 ELM 的 AUC 和准确率最高，在测试集和验证集的 AUC 和准确率分别为 97.8%和 92.8%、95%和 90%。
总之，我们观察到机器学习方法可用于预测直肠癌患者的 nCRT 反应。预处理 LOG 滤波器和 EL 模型可以改善预测过程。
