Long-Term Response Classification and Prediction of Atrial Fibrillation Ablation Using Non-Invasive 3D-Rendered Electromechanical Cycle Length Mapping

Effective treatment of atrial fibrillation (AF) remains a clinical challenge due to its complex electrophysiological patterns. This study explores the feasibility of Electromechanical Cycle Length Mapping (ECLM), a noninvasive, high-framerate 2D-ultrasound imaging technique, to map and quantify AF e...

Full description

Saved in:
Bibliographic Details
Main Authors: Tourni, Melina, Han, Seungyeon Julia, Kucinski, Mary, Biviano, Angelo, Konofagou, Elisa
Format: Conference Proceeding
Language:English
Subjects:
Arrhythmia
arrhythmia ablation
Atrial fibrillation
cycle length
echocardiography
Measurement
Patient monitoring
response prediction
Sociology
Support vector machines
Visualization
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Effective treatment of atrial fibrillation (AF) remains a clinical challenge due to its complex electrophysiological patterns. This study explores the feasibility of Electromechanical Cycle Length Mapping (ECLM), a noninvasive, high-framerate 2D-ultrasound imaging technique, to map and quantify AF electromechanical cycle length (CL) activity and inform on long-term ablation procedure response. 16 AF and six Atrial Flutter subjects were imaged pre- and post-successful first-time cardiac ablation. Pre-ablation ECLM maps identified the location of arrhythmic triggers in 100% of the subjects and showcased their reduction post-ablation. Patient monitoring revealed five AF subjects reverted to arrhythmia within the first 3-months. Pre-ablation CL dispersion values were an independent predictor of the 3-month AF recurrence (multivariate logistic regression AUC of 0.78). In addition, a Support Vector Machine (SVM) classifier trained on the ECLM-derived metrics distinguished 3-month responders (AUC of 0.89). These findings support the key role ECLM could play in immediate atrial arrhythmia evaluation and treatment response prediction.
ISSN:1945-8452
DOI:10.1109/ISBI56570.2024.10635652