AI–Enhanced Electrocardiography as a Digital Biomarker Platform in Cardiovascular Medicine

In recent years, Artificial Intelligence (AI) applied to the electrocardiogram (ECG) has shown increasing potential in improving diagnosis, prognostic stratification, and cardiovascular screening through the identification of electrophysiological patterns not detectable with conventional interpretation. This narrative review, incorporating elements of a scoping review, summarizes the main available evidence on the use of AI-ECG across the entire continuum of cardiovascular disease, including arrhythmia detection, early identification of structural heart disease, decision support in acute coronary syndromes, prediction of clinical outcomes, and applications in population screening using wearable devices. Deep learning models applied to both the standard 12-lead ECG and simplified re-cordings have demonstrated high diagnostic performance in identifying atrial fibrillation (AF), left ventricle (LV) dysfunction, hypertrophic cardiomyopathy (HCM), cardiac amyloidosis, and heart failure with preserved ejection fraction (HFpEF), as well as a potential role in cardiovascular risk stratification and in the identification of systemic digital biomarkers. Despite these promising results, the clinical adoption of AI-ECG is still limited by the need for prospective multicenter validation and by challenges related to model interpretability and their integration into clinical workflows. Overall, AI-ECG represents an emerging diagnostic platform with potential applications in predictive, preventive, and personalized cardiology.