Directed Network Mapping Approach to Rotor Localization in Atrial Fibrillation Simulation.

Area of Science:

• Electrophysiology
• Computational modeling
• Network science

Background:

• Catheter ablation for atrial fibrillation (AF) has suboptimal outcomes, especially in persistent AF.
• Current ablation targets are often anatomic, not patient-specific functional targets.
• Identifying precise ablation targets remains a significant challenge.

Purpose of the Study:

• To investigate a novel directed network approach for automatic detection of arrhythmia mechanisms.
• To guide catheter ablation strategies using patient-specific functional targets.
• To improve AF ablation outcomes by identifying stable rotor locations.

Main Methods:

• Generated directed networks from unipolar electrograms (EGMs) recorded by catheters.
• Used cross-covariance time-delay estimation to define network edges.
• Analyzed EGMs in short time frames to capture complex wave propagation patterns.
• Simulated AF using a detailed 3D computational model of human atria.

Main Results:

• Successfully identified stable rotor locations by detecting repeating cycles in consecutive networks.
• Located a stable rotor by detecting a cycle between 3 nodes in 19 out of 58 networks.
• Confirmed rotor location with 10 consecutive time frames, demonstrating algorithm's stability.

Conclusions:

• Directed network analysis of EGMs can automatically detect critical arrhythmia drivers.
• This approach offers a promising method for identifying patient-specific functional targets for AF ablation.
• The findings support the potential of this novel technique to guide more effective ablation strategies.