CT radiomics to differentiate between Wilms tumor and clear cell sarcoma of the kidney in children

BMC Medical Imaging

Table 3 Evaluation parameters of the radiomics model

Parameters	Training set	fivefold CV in the training set	Test set
AUC (95% CI)	0.889 (0.811–0.967)	0.863 (0.774–0.952)	0.792 (0.616–0.968)
Accuracy	0.864	0.852	0.857
MCC	0.662	0.627	0.634
Sensitivity	0.739	0.696	0.600
Specificity	0.914	0.914	0.960
PPV	0.773	0.762	0.857
NPV	0.898	0.883	0.857
Youden Index	0.653	0.609	0.560
F1 score	0.756	0.728	0.706

AUC Area under the curve, CI Confidence interval, MCC Matthews correlation coefficient, PPV Positive prediction value, NPV Negative prediction value, CV Cross-validation

ISSN: 1471-2342