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

Figure 5

Using Kaplan-Meier analysis to verify the performance of radiomics signature. The response ability of GBM patients to radiotherapy was successfully divided into the high-risk group (radiotherapy resistance group, ) and low-risk group (radiotherapy effective group, ) according to the prediction results of radiomics signature. There were significant differences in TCIA (a) and test (b) datasets between the high-risk group and the low-risk group.

(a) Prediction of radiomics signature in TCIA dataset

(b) Prediction of radiomics signature in test dataset