Abstract
AIMS: Ablation of typical atrial flutter (AFL) within an interventional cardiac magnetic resonance (iCMR) is a novel treatment modality. This study aims to describe the segmental kinetics of the cavotricuspid isthmus (CTI) during iCMR-guided ablation for AFL, to evaluate the impact of CTI dynamics on procedural time, to assess the utility of machine learning (ML) for clustering patient profiles based on CTI kinetics, and to identify clinical factors influencing CTI kinetics. METHODS AND RESULTS: A cohort of 32 patients underwent first-time iCMR-guided CTI ablation while in sinus rhythm, of whom 15 (47%) underwent a successful electrical cardioversion (EC) within 12 h before the procedure. CTI delineation and measurements were retrospectively performed using TOMTEC-ARENA™ software. Normalized elongation (NE) was defined as the ratio between CTI elongation and CTI length during right atrial systole. Unsupervised ML (K-means clustering) was used for patients' classification.Segmental analysis revealed greater displacements for CTI segments near the tricuspid valve compared with those near the Eustachian valve. K-means clustering identified three patient groups: low, intermediate, and high NE. Prior EC was significantly associated with low NE (P < 0.05), suggesting myocardial stunning. Hypokinetic CTIs were more prevalent among patients with dyslipidaemia, smoking history, and elevated BMI. CONCLUSION: This study provides the first detailed description of segmental CTI dynamics during iCMR-guided AFL ablation. NE emerged as a valuable metric for characterizing CTI kinetics. A clinical profile including a history of EC, smoking status, elevated BMI, and dyslipidaemia, was linked to reduced CTI kinetics suggestive of right atrial cardiomyopathy.