Table 2.
Learning curve and quantitative evaluation of model performance at group level, including AUC, sensitivity, specificity and F1 score.
| # Training Recordings |
AUC (median: range) |
Sensitivity (median: range) |
Specificity (median: range) |
F1 Score (median: range) |
|---|---|---|---|---|
| 5 | 86.71: 64.72–99.20% |
82.08: 56.00–100% |
77.88: 36.81–96.41% |
84.27: 52.85–97.82% |
| 10 | 84.11: 57.63–98.47% |
80.92: 41.29–99.42% |
73.01: 47.74–97.58% |
83.32: 60.53–98.45% |
| 15 | 86.69: 63.12–99.26% |
80.20: 56.74–99.29% |
79.94: 61.45–96.39% |
84.92: 73.36–97.27% |
| 20 | 87.88: 65.14–99.30% |
80.56: 47.25–100% |
80.22: 57.41–95.96% |
85.82: 71.19–96.97% |
| 25 | 87.74: 63.73–99.27% |
81.92: 51.91–100% |
81.44: 61.74–95.97% |
87.31: 73.93–96.97% |
| 30 | 87.87: 63.07–99.28% |
82.64: 50.08–100% |
80.34: 59.49–96.08% |
87.38: 72.44–97.69% |