Automatic Extraction of Electrocardiographic F-waves in Atrial Fibrillation

Atrial fibrillation (AF), the most prevalent sustained cardiac arrhythmia, presents intricate diagnostic and therapeutic hurdles. Central to unraveling AF complexities lies the analysis of fibrillatory waves (F-waves) within electrocardiographic (ECG) signals. These dynamic F-waves, evolving in characteristics as AF progresses, play a pivotal role in guiding treatment strategies. This thesis rigorously explores F-wave extraction methodologies, crucial in delineating the stage and severity of AF, influencing treatment decisions profoundly. The evolving techniques, from conventional signal processing to sophisticated computational methods, underscore the need for precision in F-wave analysis. As AF advances, F-wave characteristics—such as amplitude and frequency—hold significant implications for treatment outcomes. Success rates of interventions like cardioversion or the recurrence of AF post-catheter ablation are intricately linked to these F-wave features. Refinement in F-wave analysis thus holds the promise of enhancing treatment precision and improving patient outcomes in AF management. The investigation into extraction algorithms—Average Beat Subtraction (ABS) and Principal Component Analysis (PCA) on real and simulated datasets highlights the challenges in accurately isolating F-waves amidst ECG complexities. ABS demonstrates stability but faces limitations in shorter segments, while PCA offers consistent performance across varied signal durations. This study emphasizes the critical need for comprehensive datasets and nuanced algorithms to elevate diagnostic accuracy, laying a foundation for advancements in AF management. By elucidating the significance of F-wave analysis in treatment tailoring, this research aims to contribute to refining therapeutic strategies for improved patient care in AF.