Table 4.

Predictive performance of clinic-radiological model, radiomics signature, and radiomics nomogram.

Model	Radiomics nomogram		Radiomics signature		Clinic-radiological model
	Training	Testing	Training	Testing	Training	Testing
AUC (95% CI)	0.891 (0.832-0.962)	0.843 (0.718-0.975)	0.833 (0.751-0.925)	0.804 (0.667-0.958)	0.809 (0.715-0.897)	0.722 (0.546-0.894)
Accuracy	0.822 (0.727-0.895)	0.811 (0.649-0.920)	0.778 (0.678-0.859)	0.783 (0.618-0.902)	0.756 (0.654-0.840)	0.703 (0.530-0.841)
Sensitivity	0.796	0.727	0.854	0.850	0.679	0.636
Specificity	0.848	0.933	0.691	0.706	0.882	0.801
PPV	0.833	0.941	0.759	0.773	0.905	0.824
NPV	0.813	0.709	0.806	0.804	0.625	0.621