A method for quantifying atrial fibrillation organization based on wave-morphology similarity

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Atrial fibrillation (AF) is a complex arrhythmia characterized by disorganized atrial electrical activity.
• Quantifying the organization of electrograms during AF is crucial for understanding its mechanisms and developing effective treatments.

Purpose of the Study:

• To introduce a new algorithm for quantifying the organization of bipolar electrograms during human atrial fibrillation (AF).
• To assess the algorithm's ability to differentiate AF complexity and its potential clinical utility.

Main Methods:

• The algorithm compares pairs of local activation waves (LAWs) to estimate morphological similarity.
• A regularity index (rho) is calculated, measuring the repetitiveness of detected activations over time.
• Data from multipolar basket catheter recordings during AF and atrial flutter were analyzed.

Main Results:

• The index demonstrated maximum regularity (rho = 1) during atrial flutter and decreased significantly with increasing AF complexity (Type I, II, III).
• A classification scheme based on minimum distance analysis achieved 85.5% accuracy in distinguishing AF episodes.
• The algorithm accurately discriminated AF types even with limited signal data (five LAWs).

Conclusions:

• The developed algorithm provides a reliable measure of atrial electrogram organization during AF.
• It can effectively differentiate AF complexity and detect transient changes, aiding in understanding AF mechanisms.
• This tool holds potential for improving the clinical diagnosis and treatment of AF.