Table 4 Comparisons of models in the training, testing and validation datasets
Different models | Training cohort (n = 72) | Testing cohort (n = 32) | Validation cohort (n = 26) | ||||||
|---|---|---|---|---|---|---|---|---|---|
Sensitivity | Specificity | AUC (95% CI) | Sensitivity | Specificity | AUC (95% CI) | Sensitivity | Specificity | AUC (95% CI) | |
Clinical | 91.2% | 73.3% | 0.839(0.718,0.942) | 76.9% | 50.0% | 0.750(0.563,0.917) | 90.5% | 60.0% | 0.724(0.515,0.879) |
T-RO | 80.7% | 66.7% | 0.784(0.647,0.901) | 61.5% | 83.3% | 0.763(0.607,0.900) | 76.2% | 60.0% | 0.600(0.391,0.785) |
P-RO | 86.0% | 53.3% | 0.740(0.589,0.873) | 80.8% | 33.3% | 0.756(0.568,0.929) | 61.9% | 80.0% | 0.629(0.418,0.808) |
TP-RO | 87.7% | 73.3% | 0.836(0.698,0.951) | 69.2% | 50.0% | 0.731(0.552,0.885) | 76.2% | 60.0% | 0.686(0.475,0.852) |
C-RO | 91.2% | 80.0% | 0.896(0.800,0.969) | 73.1% | 83.3% | 0.801(0.650,0.923) | 85.7% | 80.0% | 0.790(0.587,0.924) |