Artificial intelligence in cardiovascular electrophysiology: systematic review 2010-2025

Jorge E. Marín-Velázquez, Universidad CES; Servicio de Electrofisiología, Clínica CES; Servicio de Electrofisiología, Clínica Las Américas; Servicio de Electrofisiología, PUL.SOS; Medellín, Colombia
Mauricio Duque-Ramírez, Universidad CES; Servicio de Electrofisiología, PUL.SOS; Medellín, Colombia
María A. Mesa-Maya, Universidad CES; Servicio de Cardiología, Clínica CES; Servicio de Cardiología, Clínica Las Vegas; Colombia
Julián M. Aristizábal-Aristizábal, Universidad CES; Servicio de Electrofisiología, Clínica Las Américas; Servicio de Electrofisiología, PUL.SOS; Servicio de Electrofisiología, Clínica Las Vegas; Servicio de Electrofisiología, Clínica El Rosario; Servicio de lectrofisiología Hospital General de Medellín; Colombia
Oriana Bastidas-Ayala, Universidad CES; Servicio de Electrofisiología, Clínica Las Américas; Servicio de Electrofisiología, PUL.SOS; Servicio de Electrofisiología, Clínica Las Vegas; Servicio de Electrofisiología, Clínica El Rosario; Servicio de Electrofisiología, Hospital Pablo Tobón Uribe; Medellín, Colombia
César D. Niño-Pulido, Universidad CES; Servicio de Electrofisiología, Clínica Las Américas; Servicio de Electrofisiología, PUL.SOS; Servicio de Electrofisiología, Hospital Pablo Tobón Uribe; Servicio de Electrofisiología, Clínica SOMER INCARE. Medellín, Colombia
Juan C. Díaz-Martínez, Universidad CES; Servicio de Electrofisiología, Clínica Las Américas; Servicio de Electrofisiología, PUL.SOS; Servicio de Electrofisiología, Clínica Las Vegas; Servicio de Electrofisiología, Clínica El Rosario; Servicio de lectrofisiología Hospital General de Medellín, Medellín, Colombia

Introduction: Artificial intelligence (AI) has significantly transformed various aspects of cardiovascular electrophysiology (EP) over the past decade. This systematic review evaluates its clinical applications between 2010 and 2025. Objective: To synthesize the evidence on AI use in five key areas of EP: electroanatomic mapping and ablation, arrhythmia prediction, catheter ablation procedures, ECG/Holter interpretation, and integration with 3D systems. Methods: A systematic review was conducted following PRISMA guidelines. Databases searched included PubMed, Scopus, Embase, and grey literature. Studies in adult humans with practical AI applications in EP were included. Results: The review found evidence that AI enhances arrhythmia detection (e.g., atrial fibrillation, ventricular tachyarrhythmias), improves event prediction accuracy over traditional risk scores, and supports therapeutic procedures (e.g., VX1 in persistent AF). AI also showed clinical-grade performance in automated ECG and Holter analysis. Conclusions: AI represents a major advancement in EP, with some tools already in clinical use. Despite ongoing challenges – such as large-scale validation and regulatory frameworks – AI holds substantial promise for enhancing the diagnosis, risk stratification, and treatment of cardiac arrhythmias.

Keywords: Artificial intelligence. Cardiac electrophysiology. Cardiac arrhythmias. Machine learning. Catheter ablation. Electrocardiography.