TABLE 4. The AUROC (Sensitivity) of the Classifier on Training and Test Sets of Data for Different Pre- and Post-Alarm Windows, for Individual Feature-Families and the Combined Feature-Pool With the Specificity Fixed at 0.98. Fot the Training Set, the Performace Results are the Mean Cross-Validated Estimate.
| Pre-, post-alarm window (min) | Metadata | Correlation | Alarm counts | BR | HR | HRV | Alarm category | SpO2 | Combination | Total no. of combined features |
|---|---|---|---|---|---|---|---|---|---|---|
| Training set | ||||||||||
| 3,3 | 0.60 (0.05) | 0.57 (0.05) | 0.65 (0.07) | 0.65 (0.07) | 0.66 (0.09) | 0.68 (0.09) | 0.70 (0.06) | 0.78 (0.17) | 0.80 (0.21) | 23 |
| 3,2 | 0.59 (0.04) | 0.58 (0.06) | 0.66 (0.07) | 0.67 (0.08) | 0.69 (0.12) | 0.69 (0.09) | 0.72 (0.07) | 0.79 (0.19) | 0.83 (0.22) | 22 |
| 3,1 | 0.59 (0.04) | 0.61 (0.07) | 0.67 (0.07) | 0.70 (0.09) | 0.72 (0.14) | 0.73 (0.11) | 0.74 (0.08) | 0.82 (0.22) | 0.86 (0.29) | 23 |
| 2,1 | 0.59 (0.04) | 0.60 (0.07) | 0.67 (0.07) | 0.70 (0.09) | 0.72 (0.13) | 0.73 (0.11) | 0.74 (0.08) | 0.82 (0.22) | 0.87 (0.29) | 21 |
| 1,1 | 0.59 (0.04) | 0.60 (0.07) | 0.67 (0.07) | 0.71 (0.09) | 0.73 (0.13) | 0.73 (0.12) | 0.74 (0.08) | 0.82 (0.22) | 0.86 (0.29) | 23 |
| Test set | ||||||||||
| 3,3 | 0.60 (0.05) | 0.57 (0.05) | 0.65 (0.07) | 0.64 (0.06) | 0.66 (0.09) | 0.68 (0.09) | 0.70 (0.06) | 0.78 (0.17) | 0.80 (0.21) | 23 |
| 3,2 | 0.59 (0.04) | 0.58 (0.06) | 0.66 (0.08) | 0.66 (0.07) | 0.68 (0.11) | 0.69 (0.09) | 0.72 (0.07) | 0.79 (0.19) | 0.83 (0.22) | 22 |
| 3,1 | 0.59 (0.04) | 0.61 (0.07) | 0.67 (0.07) | 0.69 (0.08) | 0.72 (0.13) | 0.73 (0.10) | 0.74 (0.08) | 0.82 (0.22) | 0.86 (0.29) | 23 |
| 2,1 | 0.59 (0.04) | 0.60 (0.07) | 0.67 (0.07) | 0.70 (0.09) | 0.72 (0.13) | 0.73 (0.11) | 0.74 (0.08) | 0.82 (0.22) | 0.87 (0.29) | 21 |
| 1,1 | 0.59 (0.04) | 0.60 (0.07) | 0.67 (0.07) | 0.71 (0.09) | 0.73 (0.13) | 0.73 (0.12) | 0.74 (0.08) | 0.82 (0.22) | 0.86 (0.29) | 23 |