TABLE 2.
Deep learning model performance mean metrics (95% CI).
| Hospital 1 | Hospital 2 | |||
|---|---|---|---|---|
| Performance Metric | Creatinine | Bilirubin | Creatinine | Bilirubin |
| AUROC | 0.99 (0.98, 1) | 0.97 (0.95, 0.99) | 0.76 (0.70, 0.82) | 0.72 (0.68, 0.76) |
| F1 Score | 0.99 (0.98, 1) | 0.66 (0.65, 0.67) | 0.59 (0.54, 0.64) | 0.24 (0.14, 0.33) |
| Sensitivity | 0.99 (0.98, 1) | 0.99 (0.98, 0.99) | 0.60 (0.55, 0.64) | 0.54 (0.52, 0.56) |
| Specificity | 0.99 (0.98, 1) | 0.91 (0.89, 0.93) | 0.98 (0.96, 0.99) | 0.90 (0.87, 0.94) |
| PPV | 0.99 (0.98, 1) | 0.5 (0.5, 0.5) | 0.59 (0.47, 0.71) | 0.24 (0.23, 0.25) |
| NPV | 0.99 (0.98, 1) | 0.99 (0.99, 0.99) | 0.99 (0.98, 1) | 0.92 (0.91, 0.94) |
| Cohen's Kappa | 0.99 (0.98, 1) | 0.65 (0.60, 0.69) | 0.57 (0.5, 0.64) | 0.20 (0.10, 0.31) |
| FNR | 0.01 (0, 0.02) | 0.00 (0.00, 0.00) | 0.31 (0.27, 0.36) | 0.37 (0.28, 0.45) |
Abbreviations: AUROC, area under receiver operator characteristic curve; FNR, false negative rate; NPV, negative predictive power; PPV, positive predictive power.