Artificial Intelligence-Augmented Electrocardiogram Detection of Left Ventricular Systolic Dysfunction in the General Population

The goal of this study was to validate an AI-enabled ECG algorithm for the detection of preclinical left ventricular systolic dysfunction (LVSD) in a large community-based cohort. We found that for detection of LVSD in the total population, the area under the curve (AUC) for the AI-ECG was 0.97, and in the high-risk subgroup, the AUC was 0.97 as well. Overall, the study showed that AI-enabled ECGs can identify preclinical LVSD in the community and requires further study as a screening tool for preclinical LVSD.