Table 3. Performance comparison between the optimal radiomics model (radiomics-A), the optimal DLR model (DLR-A), and the model based on CT findings (56 patients from Hospital I).

Model	ACC	SEN	SPC	AUC	P
Radiomics-A	0.75	0.70	0.76	0.74	0.020
DLR-A	0.71	0.90	0.67	0.80	0.003
CT findings	0.63	0.50	0.65	0.53	0.748

A P value indicates the significance level of the comparison between an AUC with that of a random case (AUC =0.5). DLR, deep learning radiomics; A, arterial; CT, computed tomography; ACC, accuracy; SEN, sensitivity; SPC, specificity; AUC, area under the curve.