Table 4.
Performances of different models in predicting the postoperative visual recovery of pituitary adenoma patients in the development set and independent test set
| AUC [95% CI] | Accuracy | Sensitivity | Specificity | F1-score | PPV | NPV | |
|---|---|---|---|---|---|---|---|
| Development set | |||||||
| Age | 0.641 [0.511—0.771] | 0.703 | 0.803 | 0.500 | 0.784 | 0.766 | 0.556 |
| Delta-radiomic model | 0.821 [0.729—0.914] | 0.747 | 0.721 | 0.800 | 0.793 | 0.880 | 0.585 |
| Combined model | 0.841 [0.750—0.932] | 0.714 | 0.607 | 0.933 | 0.740 | 0.949 | 0.538 |
| Independent test set | |||||||
| Age | 0.655 [0.506—0.805] | 0.744 | 0.846 | 0.538 | 0.815 | 0.786 | 0.636 |
| Delta-radiomic model | 0.811 [0.692—0.930] | 0.846 | 0.923 | 0.692 | 0.889 | 0.857 | 0.818 |
| Combined model | 0.840 [0.733—0.947] | 0.769 | 0.692 | 0.923 | 0.800 | 0.947 | 0.600 |
AUC area under the receiver operating characteristic curve, CI confidence interval, PPV positive predictive value, NPV negative predictive value