Table 2. Diagnostic performance of model.
Model | Set | AUC | Accuracy | Sensitivity | Specificity | F1 |
---|---|---|---|---|---|---|
Intra-radiomics | Training | 0.797 | 0.785 | 0.534 | 0.905 | 0.73 |
Test | 0.780 | 0.754 | 0.526 | 0.868 | 0.667 | |
Intra + peri-radiomics | Training | 0.910 | 0.868 | 0.721 | 0.938 | 0.779 |
Test | 0.840 | 0.745 | 0.500 | 0.868 | 0.567 | |
Intra + peri + clinical | Training | 0.950 | 0.913 | 0.791 | 0.972 | 0.855 |
Test | 0.960 | 0.921 | 0.816 | 0.974 | 0.873 |
AUC, area under the curve; intra, intratumoral features; peri, peritumoral features.