Dynamic approximate entropy electroanatomic maps detect rotors in a simulated atrial fibrillation model

Area of Science:

• Computational cardiology
• Non-linear dynamics in physiology
• Cardiac electrophysiology

Background:

• Rotors are implicated as drivers of atrial fibrillation (AF).
• Complex fractionated atrial electrograms (CFAEs) are found near rotor tips.
• Current methods for identifying ablation targets based on CFAE time intervals are controversial.

Purpose of the Study:

• To investigate the relationship between approximate entropy and electrogram fractionation.
• To develop a novel tool for rotor mapping using approximate entropy.
• To determine if approximate entropy can locate rotor tips in AF.

Main Methods:

• Simulated two episodes of chronic AF in a 3D human atrial model.
• Calculated dynamic approximate entropy maps from simulated unipolar electrograms.
• Optimized approximate entropy using real multi-center databases of fractionated electrograms.
• Validated rotor detection using bipolar electrograms near simulated rotors.

Main Results:

• Approximate entropy values positively correlated with electrogram fractionation levels.
• Dynamic approximate entropy maps successfully localized stable and meandering rotor tips.
• Optimized approximate entropy detected rotors using bipolar electrograms.

Conclusions:

• High approximate entropy values indicate significant fractionation and can locate AF rotor tips.
• Dynamic approximate entropy mapping shows promise as a tool for AF rotor mapping.
• This approach may improve the identification of ablation targets in AF.