Noise and spatial-resolution effect of electrode array on rotor tip location during atrial fibrillation: A simulation study

Atrial fibrillation (AF) is the most common arrhythmia in clinical practice. Recently, a mechanism for maintaining the AF which consists of one or more rotors activating the tissue at high frequency has been proposed. Ablation is one of the treatments for AF whose effectiveness could depend on the localization of the rotor tip. In a previous research, the approximate entropy (ApEn) was proposed as an effective rotor tip detection system. However, this system was validated with an array of 22500 electrodes providing one noise-free simulated electrogram per electrode, which can be considered as a system distant to the reality. In this work, three analyses from simulated electrograms obtained from 2D model of human atrial tissue under chronic AF conditions for rotor tip detection are presented. The first one with noise-free, the second one with noise of real electrograms (NoRE), and the third one is done by applying the Kalman filter (KF) on the simulated electrograms with NoRE. In each case, the effect of the dimension of electrodes array is analyzed with relationships of equidistant reduction rows-column from 2-2 to 6-6. This comparison showed that the noise has a greater negative effect than the one generated by the decrease of resolution in electrodes array, and despite of the acceptable Kalman filter's performance for increasing the SNR significantly, the accuracy is lower than the results obtained with electrograms with NoRE but adequate for locating the rotor tip.