Table 2 Performance metrics of three models
Accuracy | AUC (95%CI) | Sensitivity | Specificity | |
|---|---|---|---|---|
Dosiomics model | ||||
Training | 0.9643 | 0.9986 (0.9962–1.000) | 0.9474 | 1.000 |
Test | 0.8214 | 0.8462 (0.7156–0.9767) | 1.000 | 0.9130 |
DLR model | ||||
Training | 0.9464 | 0.9992 (0.9978–1.000) | 1.000 | 0.9744 |
Test | 0.7857 | 0.8750 (0.7508–0.9992) | 1.000 | 0.8095 |
Hybrid model | ||||
Training | 0.9642 | 0.9992 (0.9977–1.000) | 1.000 | 0.9867 |
Test | 0.8571 | 0.9000 (0.7895–1.000) | 1.000 | 0.8750 |