A triple-classification for differentiating renal oncocytoma from renal cell carcinoma subtypes and CK7 expression evaluation: a radiomics analysis

Table 2 The diagnostic performance of the radiomic models in the training set (n = 123)

Classifier evaluation	CMP			NP			Combination
Classifier evaluation	RO	chRCC	ccRCC	RO	chRCC	ccRCC	RO	chRCC	ccRCC
Average AUC	0.864	0.898	0.871	0.853	0.865	0.884	0.928	0.955	0.939
95% CI	0.792, 0.937	0.828, 0.968	0.780, 0.962	0.724, 0.933	0.797, 0.933	0.800, 0.969	0.838, 0.997	0.913, 0.996	0.880, 0.997
Balanced Accuracy	0.822	0.918	0.863	0.845	0.832	0.900	0.928	0.940	0.943
Average Sensitivity	0.844	0.950	0.775	0.789	0.850	0.825	0.956	0.950	0.900
Average Specificity	0.800	0.886	0.971	0.900	0.814	0.957	0.900	0.929	0.986

AUC, the area under the curve. CI, confidence interval. CMP, corticomedullary phase. NP, nephrographic phase

ISSN: 1471-2490