Skip to main content
. 2021 Aug 11;15:721268. doi: 10.3389/fnins.2021.721268

TABLE 5.

The p-values of the DeLong test of the statistical comparison of the ROC curves in all datasets.

Training dataset Temporal validation dataset External validation dataset
R-model vs. M-model 0.458 0.211 0.457
M-model vs. RM-model 0.041 0.005 0.224
R-model vs. RM-model 0.176 0.115 0.559
R-model vs. CM-model 0.743 0.685 0.603
M-model vs. CM-model 0.515 0.212 0.106
R-model vs. CRM-model 0.078 0.041 0.407
M-model vs. CRM-model 0.018 0.002 0.152
RM-model vs. CRM-model 0.176 0.096 0.378
CM-model vs. CRM-model 0.038 0.018 0.993

R-model, radiomics model; M-model, morphological model; RM-model, radiomics-morphological model; CM-model, clinical-morphological model; CRM-model, clinical-radiomics-morphological model.