Machine Learning Based on a Multiparametric and Multiregional Radiomics Signature Predicts Radiotherapeutic Response in Patients with Glioblastoma

Figure 1

A flowchart describing the radiomics method for prediction of radiotherapy response. (1) Preprocess the original image; (2) delineate the subregion ROI by GLISTR; (3) feature extraction using pyradiomics; (4) 31-gene signature were used to predict the result of the corresponding data, and a label was generated. (5) Feature selection by the Boruta algorithm; (6) modeling by a variety of machine learning methods, ROC curve, and AUC evaluation model; (7) building a prediction model by combining radiomics signature features and clinical features, finally displaying the OS prediction results by nomogram.