Table 6 Models trained on ECG, art line, and EHR data, 12-h gap.
From: Continuous sepsis trajectory prediction using tensor-reduced physiological signals
Model | Rank | F1 score | Recall | Specificity | AUROC |
|---|---|---|---|---|---|
LUCCK | 1 | 0.57 (0.09) | 0.82 (0.09) | 0.78 (0.09) | 0.82 (0.06) |
2 | 0.56 (0.09) | 0.80 (0.09) | 0.79 (0.08) | 0.82 (0.06) | |
3 | 0.55 (0.09) | 0.81 (0.08) | 0.78 (0.08) | 0.82 (0.06) | |
4 | 0.56 (0.09) | 0.83 (0.09) | 0.78 (0.08) | 0.83 (0.06) | |
None | 0.55 (0.08) | 0.81 (0.10) | 0.78 (0.08) | 0.80 (0.06) | |
RF | 1 | 0.58 (0.10) | 0.82 (0.10) | 0.79 (0.08) | 0.84 (0.05) |
2 | 0.57 (0.08) | 0.80 (0.09) | 0.80 (0.07) | 0.84 (0.06) | |
3 | 0.56 (0.08) | 0.83 (0.10) | 0.78 (0.07) | 0.84 (0.05) | |
4 | 0.57 (0.09) | 0.83 (0.09) | 0.78 (0.09) | 0.84 (0.06) | |
None | 0.57 (0.09) | 0.85 (0.09) | 0.78 (0.09) | 0.83 (0.06) | |
SVM | 1 | 0.43 (0.09) | 0.71 (0.14) | 0.68 (0.09) | 0.67 (0.10) |
2 | 0.39 (0.10) | 0.67 (0.15) | 0.65 (0.13) | 0.61 (0.13) | |
3 | 0.37 (0.09) | 0.64 (0.16) | 0.64 (0.12) | 0.58 (0.13) | |
4 | 0.35 (0.09) | 0.60 (0.15) | 0.63 (0.15) | 0.53 (0.11) | |
None | 0.53 (0.09) | 0.80 (0.10) | 0.75 (0.09) | 0.79 (0.07) |
