CT differentiation of the oncocytoma and renal cell carcinoma based on peripheral tumor parenchyma and central hypodense area characterisation

Table 4 Diagnostic performance of predictive models for differentiation of ROs and RCCs

Model parameter	Model−1	Model−1^*	Model−1^*_validation	Model−2	Model−2^*	Model−2^*_validation
Sensitivity	0.783(18/23)	0.826(19/23)	0.875(7/8)	0.783(18/23)	0.826(19/23)	0.750(6/8)
Specificity	0.953(81/85)	0.965(82/85)	0.829(29/35)	0.833(20/24)	0.875(21/24)	0.900(9/10)
PPV	0.818(18/22)	0.864(19/22)	0.538(7/13)	0.818(18/22)	0.864(19/22)	0.857(6/7)
NPV	0.942(81/86)	0.953(82/86)	0.967(29/30)	0.800(20/25)	0.840(21/25)	0.818(9/11)
Accuracy	0.917(99/108)	0.935(101/108)	0.837(36/43)	0.809(38/47)	0.851(40/47)	0.833(40/47)
AUC	0.952	0.962	0.936	0.839	0.914	0.938
AUC(95% CI)	0.912–0.992	0.925–0.999	0.860-1.000	0.722–0.955	0.828-1.000	0.825-1.000

Model-1 and Model-2 are predictive models excluding CT features of CHA for differentiating RO from ccRCC and non-ccRCC, respectively. Model-1* and Model-2* are predictive models including CT features of CHA for differentiating RO from ccRCC and non-ccRCC, respectively
Model−1^*_validation and Model−2^*_validation are the test results of Model-1^* and Model-2^* on the validation dataset, respectively
Values are ratios of the numerator and denominator in parentheses
AUC Area under curve, CI Confidence interval, CHA Central hypodense area, ccRCC Clear cell renal cell carcinoma, NPV Negative predictive value, PPV Positive predictive value, RO Renal oncocytoma, RCC Renal cell carcinoma

ISSN: 1471-2342