Table 3 Predictive performance of the clinical model, SVM-based radiomics model and fusion model in the training, internal validation and external validation cohorts
Cohort | Model | AUC (95% CI) | SEN (%) | SPE (%) | ACC (%) |
|---|---|---|---|---|---|
Training | Clinical | 0.737 (0.669—0.797) | 71.74 | 64.42 | 67.86 |
|  | SVM-based radiomics | 0.884 (0.831—0.925) | 86.96 | 75.00 | 80.61 |
|  | Fusion | 0.916 (0.868—0.951) | 91.30 | 79.81 | 85.20 |
Internal validation | Clinical | 0.715 (0.606—0.808) | 90.00 | 50.00 | 69.05 |
|  | SVM-based radiomics | 0.860 (0.767—0.926) | 85.00 | 81.82 | 83.33 |
|  | Fusion | 0.903 (0.819—0.957) | 80.00 | 90.91 | 85.71 |
External validation | Clinical | 0.654 (0.532—0.763) | 54.55 | 76.32 | 66.20 |
|  | SVM-based radiomics | 0.841 (0.734—0.917) | 81.82 | 84.20 | 83.10 |
|  | Fusion | 0.885 (0.787—0.949) | 84.85 | 84.21 | 84.51 |