New mapping algorithm SHIM (SVD, Hilbert Identifying Method) for the extraction of phase singularities and spiral waves in cardiac arrhythmias

Y. Biton,D. Braunstein,A. Rabinovitch

DOI: https://doi.org/10.1016/j.bspc.2024.107106

IF: 5.1

2024-10-27

Biomedical Signal Processing and Control

Abstract:Background and Objective Atrial fibrillation (AF) is the most prevalent cardiac arrhythmia worldwide. The effectiveness of AF ablation is restricted primarily by inaccuracies in the voltage-map videos (VMVs) acquired through multielectrode catheters. We introduce an innovative algorithm improving map quality and pinpointing the precise locations of action potential singularities. Methods The methodology relies on three key operations: singular value decomposition (SVD), Hilbert transform, and identifying the ridge of singularities SHIM (SVD, Hilbert, Identifying Method). To check its feasibility, this approach was used to analyze record sources from two distinct origins: a) cultures of cardiac cells, and b) real-world atrial fibrillation (AF) measurements. Efficacy is exemplified through four illustrative cases, two involving cultures of cardiac cells exposing action potential spirals, and two clinical cases involving basket catheter voltage-map videos. Resolution goodness is measured by a new method of the ratio of the peak's height divided by the average width at half maximum or the width of the ridge-of-peaks cross-section. Results In the first sample, the driving rotor became evident only following the SVD procedure stage, and its stable center became more pronounced after the H and I stages. In the case of spiral splitting, SHIM extracted all the individual tiny spiral centers. In the VMVs obtained from the basket catheters, even when the image sharpness was suboptimal, distinct singularity ridges were easily discernible with peak intensities larger than 1. Conclusion The cumulative impact of the three stages of our analysis suggests that our novel method may excel in identifying problematic sources more effectively than previously utilized systems.

engineering, biomedical