Artificial Intelligence–Enhanced Electrocardiography Identifies Patients With Normal Ejection Fraction at Risk of Worse Outcomes

Summary

This Mayo Clinic paper demonstrates that AI-ECG analysis can risk-stratify patients with visually normal left ventricular ejection fraction (LVEF ≥ 55%) who are at elevated risk of worse long-term outcomes. The fundamental insight is that subclinical cardiac dysfunction — invisible to standard clinical interpretation — leaves a detectable signature in ECG waveform morphology. The AI model, trained on large-scale Mayo Clinic ECG+echocardiography data, identifies a high-risk subgroup within the clinically 'normal' LVEF range.