Atrial fibrillation (AF) initially starts in a paroxysmal form, which is mainly based on the presence of triggers. Progression of AF is accompanied by structural and electrical remodelling which can typically be described as a progressive change in electrophysiological properties of the myocardium caused by cardiovascular comorbidities and AF itself. This results in complex electrical conduction disorders, which is defined as electropathology. The aim of our study is to evaluate whether an artificial intelligence (AI)-enabled ECG algorithm (on an ECG in Sinus Rhythm) can predict electropathology in patients with AF who undergo a first PVI.