Table 2 Prediction performance on the test set. Performance metrics are given for each model and the target variables TR and MRD separately.
Performance Metrics | Radiomics Features Only Model (I) | Radiomics Features + Confounders Model (II) | Radiomics + Clinical Features Model (III) | Clinical Features Only Model (IV) | |
|---|---|---|---|---|---|
TR | AUROC (95%CI) | 0.70 (0.49–0.91) | 0.69 (0.47–0.91) | 0.69 (0.47–0.91) | 0.63 (0.41–0.85) |
Sensitivity | 0.86 | 0.75 | 0.93 | 0.93 | |
Specificity | 0.55 | 0.64 | 0.45 | 0.45 | |
F1-Score | 0.84 | 0.79 | 0.87 | 0.87 | |
MRD | AUROC (95%CI) | 0.53 (0.34–0.72) | 0.52 (0.33–0.71) | 0.54 (0.34–0.73) | 0.35 (0.17–0.53) |
Sensitivity | 0.30 | 0.35 | 0.35 | 1.00 | |
Specificity | 0.95 | 0.89 | 0.89 | 0.11 | |
F1-Score | 0.44 | 0.48 | 0.48 | 0.70 |
