Table 3.
Multivariable logistic regression results for predicting the IDH status in the study set and the test set.
| Number | Se | Sp | PPV | NPV | YI | Accuracy | AUC | |
|---|---|---|---|---|---|---|---|---|
| Study set | ||||||||
| Model 1 | 176 | 0.805 | 0.888 | 0.856 | 0.816 | 0.693 | 0.824 | 0.897 |
| Model 2 | 176 | 0.839 | 0.820 | 0.823 | 0.836 | 0.659 | 0.812 | 0.890 |
| LGG | 102 | 0.870 | 0.759 | 0.783 | 0.854 | 0.629 | 0.843 | 0.873 |
| GBM | 74 | 0.734 | 0.900 | 0.880 | 0.772 | 0.634 | 0.851 | 0.862 |
| Test set | ||||||||
| Model 1 | 40 | 0.800 | 0.800 | 0.800 | 0.800 | 0.600 | 0.750 | 0.860 |
| Model 2 | 40 | 0.700 | 0.950 | 0.933 | 0.760 | 0.650 | 0.750 | 0.893 |
| LGG | 20 | 0.833 | 0.643 | 0.700 | 0.794 | 0.476 | 0.700 | 0.762 |
| GBM | 20 | 0.857 | 0.833 | 0.837 | 0.853 | 0.690 | 0.700 | 0.833 |
Se, sensitivity; Sp, specificity; PPV, positive predictive value; NPV, negative predictive value; YI, Youden Index; AUC, area under the curve; LGG, lower-grade glioma; GBM, glioblastoma.