Skip to main content

Table 3 Performance of meningioma differentiation on the training and test sets

From: Deep learning–based automatic segmentation of meningioma from T1-weighted contrast-enhanced MRI for preoperative meningioma differentiation using radiomic features

Dataset

Sensitivity (95% CI)

Specificity

(95% CI)

Accuracy

(95% CI)

AUC

(95% CI)

Training set (automatic)

0.824

(0.738–0.910)

0.898

(0.760–0.944)

0.888

(0.753–0.921)

0.930

(0.896–0.952)

Test set (automatic)

0.778

(0.701–0.856)

0.860

(0.722–0.908)

0.848

(0.715–0.903)

0.842

(0.807–0.895)

Training set (manual)

0.941

(0.845–0.969)

0.872

(0.755–0.939)

0.881

(0.750–0.913)

0.961

(0.904–0.969)

Test set (manual)

0.778

(0.718–0.859)

0.842

(0.716–0.904)

0.833

(0.709–0.898)

0.813

(0.799–0.862)