Table 5.

ROC curve analyses for clinical imaging features, conventional radiomics, DLR and nomogram models, among training and validation sets.

Model	Accuracy	AUC (95% CI)	Sensitivity	Specificity	PPV	NPV
Training dataset
Clinical Imaging Features	0.768	0.832 (0.7753-0.8890)	0.716	0.804	0.716	0.804
Conventional Radiomics	0.774	0.861(0.8065-0.9154)	0.642	0.866	0.768	0.778
Deep Learning Radiomics	0.823	0.914 (0.8721-0.9544)	0.821	0.825	0.764	0.87
Nomogram	0.817	0.934 (0.9004-0.9680)	0.985	0.701	0.695	0.986
Validation dataset
Clinical Imaging Features	0.732	0.817 (0.7270-0.9077)	0.667	0.773	0.643	0.791
Conventional Radiomics	0.775	0.818 (0.7172-0.9183)	0.593	0.886	0.762	0.78
Deep Learning Radiomics	0.746	0.832 (0.7351-0.9290)	0.815	0.705	0.629	0.861
Nomogram	0.803	0.864 (0.7795-0.9494)	0.926	0.727	0.676	0.941