Prediction of Atrial Fibrillation Termination by Ablation with Imageless ECGI Mapping

Till Althoff1, Ines Llorente2, Ismael Hernández-Romero3, Marta Martínez Pérez4, Santiago Ros5, Jana Reventós Presmanes6, Adriana Costafreda7, �ngel Arenal8, Maite Izquierdo de Francisco9, Ivo Roca Luque1, Joaquín Osca Asensi9, Lluis Mont1, Andreu M. Climent2, Maria de la Salud Guillem Sánchez2, Felipe Atienza10
1Institut Clínic Cardiovascular, Hospital Clínic de Barcelona, Catalonia, Spain, 2Universitat Politècnica de València, 3ITACA Institute, Universitat Politècnica de València, 4COR-Group, ITACA Institute, Universitat Politècnica de València, Valencia, Spain, 5Gregorio Marañón Health Research Institute, 6Arrhythmias Department, Hospital Clínic de Barcelona, 7Corify Care, 8Hospital Gregorio Mara��n, 9Hospital Universitari i Politècnic La Fe, Valencia, Spain, 10Hospital General Universitario Gregorio Marañón (Cardiology Department)


Abstract

Background. Persistent atrial fibrillation (AF) ablation often extends beyond pulmonary vein isolation (PVI), yet substrate targeting remains empirical and lacking intra-procedural feedback. Objective To evaluate the ability of imageless electrocardiographic imaging (ECGI) to identify AF driver regions and provide real-time feedback during ablation, including prediction of termination or rhythm organization. Methods. We conducted a prospective multicenter study involving 90 patients undergoing catheter ablation for persistent AF. Imageless ECGI was performed using a 128-electrode vest placed immediately before the ablation procedure. The system allows estimating torso and cardiac geometry using a statistical shape model refined with patient-specific electrical data without using CT-based cardiac anatomy. The system metrics included the computation of stability maps that reveal which regions of both atria that exhibit the fastest activation over time. These maps quantified the percentage of time each atrial region exhibits the fastest activation rate. In a second set of 35 patients, these maps were used to guide ablation beyond PVI, targeting regions of persistent rapid activation identified prior to each catheter ablation. Results. In the observational cohort (n=90), AF terminated in sinus rhythm or organized into a stable flutter in 20 cases (i.e. 18% of cases). In 89% of these, termination occurred in regions previously identified by the stability map as exhibiting persistent high- activation rates. Across the guided cohort (n=35), the proportion of patients with acute arrhythmia modification (termination or organization) increased to 53%. Conclusion. Imageless ECGI enabled real-time visualization of rhythm evolution throughout the procedure. Operators were able to analyze the transition from AF to organized rhythms and the emergence of new driver regions, predicting arrhythmia termination. This new methodology offers immediate feedback on the electrophysiological effects of ablation and supports a more adaptive and efficient strategy for treating persistent AF.