TABLE 2.
Diagnostic performance of the deep learning system
| Training | Validation 1 | Validation 2 | Validation 2-viral# | |
| Subjects n | 709 | 226 | 161 | 121 |
| AUC 95% CI | 0.90 (0.89–0.91) | 0.87 (0.86–0.89) | 0.88 (0.86–0.90) | 0.86 (0.83–0.89) |
| Accuracy % | 81.24 | 78.32 | 80.12 | 85.00 |
| Sensitivity % | 78.93 | 80.39 | 79.35 | 79.35 |
| Specificity % | 89.93 | 76.61 | 81.16 | 71.43 |
| F1-score | 86.92 | 77.00 | 82.02 | 90.11 |
AUC: area under the receiver operating characteristic curve. #: a stratified analysis using the patients with coronavirus 2019 and viral pneumonia in validation set 2.