Abstract
The autonomic nervous system (ANS) plays an essential role in the generation and maintenance of cardiac arrhythmias. The cardiac ANS can be divided into its extrinsic and intrinsic components, with the latter being organized in an epicardial neural network of interconnecting axons and clusters of autonomic ganglia called ganglionated plexi (GPs). GP ablation has been associated with a decreased risk of atrial fibrillation (AF) recurrence, but the accurate location of GPs is required for ablation to be effective. Although GP stimulation triggers both sympathetic and parasympathetic ANS branches, a predominance of parasympathetic activity has been shown. This study aims was to develop a method to locate atrial parasympathetic innervation sites based on measurements from a grid of electrograms (EGMs). Electrophysiological models representative of non-AF, paroxysmal AF (PxAF), and persistent AF (PsAF) tissues were developed. Parasympathetic effects were modeled by increasing the concentration of the neurotransmitter acetylcholine (ACh) in randomly distributed circles across the tissue. Different circle sizes of ACh and fibrosis geometries were considered, accounting for both uniform diffuse and non-uniform diffuse fibrosis. Computational simulations were performed, from which unipolar EGMs were computed in a 16 × 1 6 electrode mesh. Different distances of the electrodes to the tissue (0.5, 1, and 2 mm) and noise levels with signal-to-noise ratio (SNR) values of 0, 5, 10, 15, and 20 dB were tested. The amplitude of the atrial EGM repolarization wave was found to be representative of the presence or absence of ACh release sites, with larger positive amplitudes indicating that the electrode was placed over an ACh region. Statistical analysis was performed to identify the optimal thresholds for the identification of ACh sites. In all non-AF, PxAF, and PsAF tissues, the repolarization amplitude rendered successful identification. The algorithm performed better in the absence of fibrosis or when fibrosis was uniformly diffuse, with a mean accuracy of 0.94 in contrast with a mean accuracy of 0.89 for non-uniform diffuse fibrotic cases. The algorithm was robust against noise and worked for the tested ranges of electrode-to-tissue distance. In conclusion, the results from this study support the feasibility to locate atrial parasympathetic innervation sites from the amplitude of repolarization wave.