Table 1.
Dataset for identifying left ventricular dysfunction (LVD) by predicting left ventricular ejection fraction (LVEF) from ECG waveforms of 92,446 unique patients from 5 hospitals at the Mount Sinai Health System
| Dataset characteristics | |||
|---|---|---|---|
| Training | Development | Test | |
| No. patients | 73,956 | 9246 | 9244 |
| Percent positive | 7.6 | 7.6 | 7.6 |
| Race (%) | |||
| White | 41.2 | 41.6 | 40.9 |
| Black | 25.1 | 25.0 | 25.4 |
| Asian | 6.6 | 6.6 | 6.4 |
| Other | 27.1 | 26.8 | 27.3 |
| Ethnicity (%) | |||
| Hispanic | 18.4 | 17.9 | 17.4 |
| Non-Hispanic | 53.1 | 53.4 | 53.9 |
| Other | 28.5 | 28.6 | 28.7 |
| Sex (%) | |||
| Male | 54.9 | 54.7 | 53.9 |
| Female | 45.1 | 45.2 | 46.0 |
| Age (mean ± SD) | 63.4 ± 15.1 | 63.2 ±14.8 | 63.5 ± 15.1 |