[The application of artificial intelligence in arrhythmology]

In recent decades, clinical practice has been founded on the principles of evidence-based medicine, where therapeutic decisions arise from the integration of clinical expertise, patient preferences, and scientific evidence derived from controlled studies and meta-analyses. The advent of artificial intelligence (AI) in health care, however, is driving a significant evolution in clinical research, owing to its ability to analyze large volumes of heterogeneous data and overcome the limitations of traditional statistical approaches. The availability of large-scale datasets, increasing computational capability, and reduced storage costs have supported the transition towards a "data-intensive" research model, progressively integrated with conventional methods. Within cardiology, arrhythmology represents one of the fields in which AI finds extensive application. The analysis of complex electrophysiological signals, data from implantable devices, advanced cardiac imaging, and clinical parameters enables the development of algorithms capable of identifying patterns not detectable by human interpretation. These tools have already demonstrated practical utility in the early diagnosis of arrhythmias, risk stratification, procedural planning and guidance for catheter ablation, prediction of response to cardiac stimulation therapies, and optimization of remote device monitoring. Among the key emerging benefits, AI promises increasingly personalized care, enabling more targeted interventions while reducing overtreatment. Furthermore, the development of "digital twins" opens the possibility of simulating patient-specific therapeutic scenarios to support complex clinical decision-making. This manuscript provides an overview of current evidence, emerging applications, and remaining challenges related to the integration of AI in arrhythmology, highlighting its potential to drive a transition towards predictive, preventive, and personalized cardiovascular medicine.