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

Figure 3

Hierarchical clustering was used to determine the expression pattern of the 31-gene signature on the samples from TCGA (Figure 3(a)). The samples in the red branch on the left side of the dendrogram are classified as Cluster0, while the samples in the blue branch on the right side are classified as Cluster1. After using Kaplan-Meier survival analysis, the prognosis of Cluster0 is different from that of Cluster1, so according to Kim et al.’s report, Cluster0 was subclassified as the radiotherapy resistance group (RR), whereas Cluster1 was a radiotherapy effective group (RE) (Figure 3(b)).

(a)

(b)