Table 3.
The comparison of the performance indices of artificial neural network (ANN) and Cox regression models for predicting mortality among hip fracture surgery patients.
| Sensitivity | Specificity | PPV | NPV | Accuracy | AUROC | |
|---|---|---|---|---|---|---|
| Training dataset (n = 7374) | ||||||
| ANN (95% CI) | 0.94 (0.91–0.98) |
0.78 (0.75–0.82) |
0.89 (0.85–0.93) |
0.82 (0.78–0.86) |
0.93 (0.90–0.96) |
0.93 (0.90–0.95) |
| Cox (95% CI) | 0.90 (0.86–0.94) |
0.67 (0.62–0.72) |
0.80 (0.74–0.86) |
0.73 (0.67–0.80) |
0.88 (0.84–0.92) |
0.89 (0.84–0.94) |
| p value | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 |
| Testing dataset (n = 1580) | ||||||
| ANN (95% CI) | 0.96 (0.92–0.99) |
0.76 (0.72–0.80) |
0.88 (0.85–0.92) |
0.84 (0.80–0.88) |
0.93 (0.89–0.97) |
0.93 (0.90–0.96) |
| Cox (95% CI) | 0.92 (0.88–0.97) |
0.64 (0.59–0.69) |
0.78 (0.72–0.84) |
0.77 (0.71–0.83) |
0.90 (0.86–0.94) |
0.88 (0.83–0.93) |
| p value | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 |
PPV = positive predictive value; NPV = negative predictive value; AUROC = area under receiver operating characteristic; CI = confidence interval.