Comparison of Artificial Intelligence–Derived Heart Age with Chronological Age Using Normal Sinus Electrocardiograms in Patients with No Evidence of Cardiac Disease
Deep learning model predicts heart age from electrocardiograms in individuals without cardiac disease, indicating a potential biomarker for biological aging.
Key Points
The study developed a deep learning model to predict heart age from electrocardiograms in healthy individuals, demonstrating strong predictive performance.
Model evaluation showed an R2 value of 0.783 and mean absolute error of 5.023 years, indicating reliability in estimating heart age.
External validation with independent ECGs confirmed robustness, achieving an R2 of 0.703 and an MAE of 5.582 years.
ECG-derived heart age could offer a reliable biomarker for biological aging and improve risk assessment strategies in clinical settings.