TABLE 4.

Performance of the radiomics, morphological, radiomics-morphological, clinical-morphological, and clinical-radiomics-morphological models.

Datasets	Models	AUC (95% CI)	ACC	SEN	SPE	PPV	NPV
Training dataset	R-model	0.822 (0.776, 0.867)	0.826	0.912	0.645	0.844	0.778
	M-model	0.798 (0.749, 0.846)	0.733	0.680	0.844	0.902	0.556
	RM-model	0.848 (0.810, 0.885)	0.795	0.788	0.809	0.897	0.644
	CM-model	0.811 (0.770, 0.853)	0.758	0.761	0.752	0.866	0.599
	CRM-model	0.856 (0.820, 0.892)	0.756	0.707	0.858	0.913	0.582
Temporal validation dataset	R-model	0.817 (0.744, 0.890)	0.800	0.928	0.653	0.755	0.887
	M-model	0.751 (0.674, 0.828)	0.690	0.590	0.806	0.778	0.630
	RM-model	0.865 (0.807, 0.924)	0.813	0.855	0.764	0.807	0.821
	CM-model	0.795 (0.723, 0.867)	0.755	0.819	0.681	0.747	0.766
	CRM-model	0.882 (0.828, 0.936)	0.832	0.928	0.722	0.794	0.897
External validation dataset	R-model	0.691 (0.567, 0.816)	0.693	0.721	0.656	0.738	0.636
	M-model	0.624 (0.490, 0.759)	0.680	0.953	0.313	0.651	0.833
	RM-model	0.721 (0.601, 0.841)	0.733	0.744	0.719	0.780	0.676
	CM-model	0.738 (0.621, 0.855)	0.747	0.860	0.594	0.740	0.760
	CRM-model	0.738 (0.618, 0.857)	0.760	0.767	0.750	0.805	0.706

R-model, radiomics model; M-model, morphological model; RM-model, radiomics-morphological model; CM-model, clinical-morphological model; CRM-model, clinical-radiomics-morphological model; AUC, area under the receiver operating curve; ACC, accuracy; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value. SEN, sensitivity; SPE, specificity.