TABLE 2.
Classification performance comparison with 95% CI.
| AUC | SE | SP | F1-Score | ACC | |
| Automated ECG feature extraction | 98.99% (96.89–100) | 96.97% (82.54–99.89) | 100% (62.97–100) | 98.46% (90–100) | 97.62% (87.44–99.99) |
| Conventional QRS morphological feature extraction | 95.62% (89.78–100) | 93.94% (78.64–98.99) | 88.89% (50.86–99.45) | 95.38% (86.62–98.86) | 92.86% (80.35–98.85) |
| Cardiologists | NA | 97.86% | 81.72% | 96.39% | 94.29% |
F1-score = 2 × Precision × recall / (precision + recall); SE, sensitivity; SP, specificity; ACC, accuracy; CI, confidence interval.