Artificial intelligence (AI) is the use of computational models to learn from electrical, anatomical and imaging data to assist or automate interpretation, prediction and decision making in arrhythmia diagnosis and treatment. This includes interpretation of signals, integration of multimodal data to support procedural decisions and to predict outcomes. When personalized to individual patients, these mechanistic approaches give rise to cardiac digital twins capable to procedural planning and hypothesis testing. In the electrophysiology (EP) lab, AI aims to reduce inter-observer variability, improve identification of the arrhythmogenic substrate by combining information from multimodality imaging and promises to streamline mapping and ablation workflows. Integration of AI with cardiac CT and cardiac MRI allows for automated segmentation, wall thickness and scar characterization and identification of conducting channels for ventricular tachycardia. Similarly, AI guided approaches for spatiotemporal dispersion and focal drivers in atrial fibrillation have demonstrated improved procedural consistency and promising clinical outcomes. This review synthesizes the current evidence on use of AI in the EP lab particularly preprocedural planning, intra procedural imaging and use of digital twins. We highlight practical workflows, representative clinical use cases and key limitations of AI.
Anand et al. (2026) studied this question.