Skip to main content
. 2020 May 28;10:680. doi: 10.3389/fonc.2020.00680

Table 2.

Diagnostic performance of binary classifiers and radiologists in the validation set.

Value (95%CI) McNemar's P-value Delong P-value
Model C Model D Reader consensus Model C vs. Reader Model D vs. Reader Model C vs. Model D
AUC 0.951 (0.919, 0.982) 0.9416 (0.914, 0.979) 0.664
Sensitivity, % 91.9 (84.7, 96.5) 90.9 (83.4, 95.8) 89.1 (76.4, 96.4) 0.375 0.219
Specificity, % 94.1 (87.6, 97.8) 94.1 (87.6, 97.8) 90.4 (79.0, 96.8) 0.549 0.754