Spiral wave classification using normalized compression distance: Towards atrial tissue spatiotemporal...

Area of Science:

• Computational electrophysiology
• Cardiac electrophysiology modeling
• Biomedical signal processing

Background:

• Atrial fibrillation (Afib) analysis requires understanding re-entrant electrical activation patterns.
• Spiral waves are key to re-entry mechanisms in cardiac tissue.
• Classifying spiral wave behavior is essential for diagnosing Afib.

Purpose of the Study:

• To develop and validate an automated method for classifying spiral wave behaviors during simulated Afib.
• To compare the effectiveness of Normalized Compressed Distance (NCD) and Normalized FFT (NFFTD) for this classification task.
• To assess the impact of catheter type, location, and recording configuration (monopolar/bipolar) on classification accuracy.

Main Methods:

• Simulated spiral wave behaviors (stable, meandering, breakup) using a cardiac electrical propagation model on a 2D grid.
• Simulated intracardiac electrograms mimicking star-shaped and circular catheters.
• Applied NCD and NFFTD techniques for classifying simulated activation patterns.

Main Results:

• NCD demonstrated superior classification performance compared to NFFTD.
• Classification accuracy showed slight variations based on catheter position and type.
• The proposed method provides theoretical validation for qualitative wavefront assessment.

Conclusions:

• Automated classification of spiral wave behavior using NCD is feasible from simulated catheter recordings.
• This approach offers a potential non-invasive method for characterizing electrophysiological activity during Afib.
• Findings suggest a qualitative assessment of wavefront patterns is possible without complex mapping.