Electroanatomical mapping based on discrimination of electrograms clusters for localization of critical sites in atrial fibrillation

Locating critical sites on the atrial surface during AF to guide the ablation procedures is an open problem. Electrogram-guided approaches have been proposed. However, electrograms (EGM) are complex and not well-described type of signals and anatomically-based pulmonary vein isolation remains been recommended as the cornerstone procedure. We introduce a method that builds an electroanatomical map to visualize the distribution of different morphological patterns of the EGM signals over the atrial surface. The proposed scheme uses EGM signals recorded with a commercial cardiac mapping. Likewise, two morphological and two non-linear features are computed from each single EGM. Patterns are discriminated using a semi-supervised clustering approach that does not need a priory definition of EGM morphologies or classes. The method was tested under two scenarios: a set of EGM signals recorded in AF patients and a set of signals obtained from 2D simulations of atrial conduction sustained by rotors. Our method was able to locate the clusters in a map of the atrial surface of each patient. These locations allow the specialist to study the distribution of critical AF sites. The method was able to locate the pivot point of the rotors in the 2D models. Our results suggest that the proposed method is a potential assisting tool for guided ablation procedures. Further clinical studies are needed to establish the relationship between clusters and arrhythmogenic substrates in AF, and to validate the usefulness of the method to locate critical conduction sites in patients. © 2018

• Fecha de publicación: 2018
• Institución: Universidad de Medellín
• Repositorio: Repositorio UDEM
• Idioma: eng
• OAI Identifier: oai:repository.udem.edu.co:11407/4854
• Acceso en línea: http://hdl.handle.net/11407/4854
• Palabra clave: Atrial fibrillation; Electroanatomical mapping; Electrograms; Semi-supervised clustering
• License: http://purl.org/coar/access_right/c_16ec