Table 3.
Performances of RS1, RS2, radiomic model, clinical factor, and predictive nomogram
| Index | Primary Cohort | Validation Cohort | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| RS1 | RS2 | radiomic model | clinical factors | predictive nomogram | senior radiologist | junior radiologist | RS1 | RS2 | radiomic model | clinical factors | predictive nomogram | senior radiologist | junior radiologist | |
| TP | 27 | 35 | 38 | 25 | 39 | 30 | 29 | 10 | 11 | 12 | 10 | 12 | 10 | 9 |
| TN | 41 | 42 | 42 | 43 | 41 | 39 | 31 | 14 | 15 | 15 | 15 | 15 | 13 | 12 |
| FN | 13 | 5 | 2 | 15 | 1 | 10 | 11 | 3 | 2 | 1 | 3 | 1 | 3 | 4 |
| FP | 5 | 4 | 4 | 3 | 5 | 7 | 15 | 2 | 1 | 1 | 7 | 1 | 3 | 4 |
| Acc | 79.07% | 89.53% | 93.02% | 79.07% | 93.02% | 80.23% | 69.77% | 82.76% | 89.66% | 93.10% | 86.21% | 93.10% | 79.31% | 72.41% |
| Sen | 0.675 | 0.875 | 0.950 | 0.625 | 0.975 | 0.750 | 0.725 | 0.769 | 0.846 | 0.923 | 0.769 | 0.923 | 0.769 | 0.692 |
| Spe | 0.891 | 0.913 | 0.913 | 0.935 | 0.891 | 0.848 | 0.673 | 0.875 | 0.938 | 0.938 | 0.938 | 0.938 | 0.813 | 0.750 |
| PPV | 0.844 | 0.897 | 0.904 | 0.893 | 0.886 | 0.811 | 0.695 | 0.833 | 0.917 | 0.923 | 0.909 | 0.923 | 0.769 | 0.692 |
| NPV | 0.759 | 0.894 | 0.955 | 0.741 | 0.976 | 0.796 | 0.738 | 0.824 | 0.882 | 0.938 | 0.833 | 0.938 | 0.813 | 0.750 |
| AUC (95%CI) | 0.837 (0.753–0.921) | 0.948 (0.903–0.994) | 0.970 (0.939–1) | 0.849 (0.769–0.929) | 0.977 (0.953–1) | 0.799 (0.716-0.884) | 0.700 (0.602-0.797) | 0.803 (0.615–0.990) | 0.938 (0.846–1) | 0.957 (0.889–1) | 0.832 (0.677–0.987) | 0.971 (0.912–1) | 0.7910.636-0.946) | 0.721 (0.551-0.892) |
NOTE. TP true positive, TN true negative, FN false negative, FP false positive, Acc accuracy, Sen sensitivity, Spe specificity, PPV positive predictive value, NPV negative predictive value, AUC area under ROC curve, CI confidence interval