Spiral waves characterization: Implications for an automated cardiodynamic tissue characterization

Area of Science:

• Cardiovascular Physiology
• Computational Biology
• Biomedical Engineering

Background:

• Spiral waves are critical phenomena in cardiac tissue, particularly during fibrillation.
• Current methods for spiral wave detection, such as high-density mapping, require specialized equipment.
• In-silico analysis often relies on comprehensive membrane potential data from entire tissues.

Purpose of the Study:

• To develop and validate a novel characterization approach for identifying spiral wave behaviors.
• To utilize intracardiac electrogram (EGM) readings from common diagnostic catheters for localized, high-resolution analysis.
• To distinguish between stationary, meandering, and break-up rotor types.

Main Methods:

• Clustering and classification algorithms applied to simulated cardiac propagation data.
• Modeling of unipolar-bipolar EGM readings using two catheter types.
• Assessment of spiral wave behavior distances using normalized compression distance (NCD) and normalized FFT distance (NFFTD).

Main Results:

• High clustering performance achieved across various EGM reading configurations.
• NCD demonstrated superior effectiveness in distinguishing spiral wave behaviors compared to NFFTD.
• Successful identification of distinct spiral activities in behaviorally heterogeneous cardiac tissue.

Conclusions:

• The study theoretically validates clustering and classification approaches for automated EGM signal analysis.
• This method provides a potential framework for mapping EGM signals to spiral wave behaviors.
• Offers a new analysis tool for understanding cardiac tissue wavefront propagation patterns.