Table 4.

Performance of the kinetic curve type, Radscore, and the nomogram models.

Model	Training cohort				Validation cohort
	AUC (95% CI)	SEN	SPEC	ACC	AUC (95% CI)	SEN	SPEC	ACC
Kinetic curve pattern	0.78 (0.69–0.85)	0.98	0.42	0.69	0.74 (0.58–0.86)	0.91	0.45	0.68
Radscore	0.86 (0.78–0.92)	0.86	0.71	0.78	0.81 (0.67–0.91)	0.68	0.91	0.79
Nomogram	0.91 (0.83–0.96)	0.82	0.86	0.84	0.86 (0.72–0.94)	0.73	0.91	0.82

AUC, area under the ROC curve; SEN, sensitivity; SPEC, specificity; ACC, accuracy.