Table 3.
Performance metrics of clinical, radiomics and combined models across datasets.
| Model | Dataset | AUC (95% CI) | Accuracy | Sensitivity | Specificity |
|---|---|---|---|---|---|
| Clinical Model | Internal Training Set | 0.705 (0.592-0.818) | 0.64 | 0.586 | 0.824 |
| Internal Validation Set | 0.691 (0.486-0.895) | 0.677 | 0.667 | 0.714 | |
| External Validation Set | 0.653 (0.450-0.857) | 0.618 | 0.593 | 0.714 | |
| Internal Training Set | 0.835 (0.733-0.937) | 0.813 | 0.828 | 0.765 | |
| Radiomics Model | Internal Validation Set | 0.833 (0.624-1.000) | 0.806 | 0.792 | 0.857 |
| External Validation Set | 0.831 (0.641-1.000) | 0.912 | 1 | 0.571 | |
| Internal Training Set | 0.896 (0.825-0.966) | 0.787 | 0.724 | 1 | |
| Combined Model | Internal Validation Set | 0.863 (0.713-1.000) | 0.71 | 0.625 | 1 |
| External Validation Set | 0.884 (0.757-1.000) | 0.735 | 0.667 | 1 |