Cardiac magnetic resonance imaging-derived atrial fibrosis in patients with pre-atrial fibrillation.
Authors
Affiliations (9)
Affiliations (9)
- Leeds Institute of Data Analytics, University of Leeds, Leeds, England, UK [email protected].
- Leeds Institute for Cardiovascular and Metabolic Medicine, University of Leeds, Leeds, UK.
- Leeds Institute of Data Analytics, University of Leeds, Leeds, England, UK.
- Department of Cardiology, Leeds Teaching Hospitals NHS Trust, Leeds, UK.
- Luliu Hatieganu University of Medicine and Pharmacy, Cluj-Napoca, Romania.
- Liverpool Heart & Chest Hospital, Liverpool, UK.
- Department of Clinical Medicine, Aalborg University, Aalborg, Denmark.
- City St George's University of London, London, UK.
- Queen Mary University of London Barts and The London School of Medicine and Dentistry, London, UK.
Abstract
Atrial fibrosis identified on cardiac magnetic resonance (CMR) imaging has been proposed as a preprocedural imaging biomarker for patient selection for rhythm control interventions in patients with atrial fibrillation (AF). Whether atrial fibrosis is present in patients considered as 'pre-AF' is unknown. We prospectively recruited 12 participants with pre-AF as defined by the Future Innovations in Novel Detection for Atrial Fibrillation (FIND-AF machine learning algorithm, without AF diagnosed during AF screening, and compared them to 25 participants with confirmed AF. All participants underwent CMR using a 3T system with left atrial fibrosis quantification and ADAS-3D left atrial image postprocessing software. Participants with pre-AF had smaller left atrial end-systolic (33.6±9.8 vs 43.0±17.0, p=0.003) and end-diastolic (16.5±8.7 vs 28.2±14.4, p=0.007) volumes, and higher left atrial ejection fraction (59.6±14.6 vs 40.7±17.5, p=0.005) than participants with AF. The extent of atrial fibrosis was not different between those with pre-AF and AF (borderzone (%) 5.2±5.0 vs 2.9±6.9, p=0.772; borderzone fibrosis (cm) 6.2±5.8 vs 6.8±10.7, p=0.927). CMR identifies atrial fibrosis before manifest AF in patients with pre-AF as defined by a machine learning algorithm.