Reconstruction of high-resolution atrial conduction heterogeneity maps from unipolar electrograms

Abstract

Understanding characteristics of atrial activation patterns at high spatiotemporal resolution is key to uncovering the mechanisms and substrate underlying arrythmias. While the temporal resolution of modern electrical mapping systems is sufficient to capture local conduction heterogeneities and dynamics through recorded electrogram morphologies, their spatial resolution remains limited by the physical constraints of electrode arrays. In this work, we propose a computational method for reconstructing high-resolution atrial tissue and conduction heterogeneity maps from unipolar electrograms, using both simulated and experimental data.

Epicardial unipolar electrograms were recorded using a 24×8 electrode array (0.5 mm diameter, 2 mm inter-electrode spacing) from the right atrial free wall during sinus rhythm with 30 kHz sampling rate. To develop and validate the reconstruction method, we generated anatomically realistic in silico datasets with comparable spatial coverage. Computer simulations were used to produce synthetic unipolar electrograms at 30 kHz temporal resolution under controlled activation and conduction conditions, providing ground-truth information on local activation times, transmembrane potentials, and spatial heterogeneity for quantitative validation.

Using simulated data, we reconstructed conduction heterogeneity maps at sub-electrode spacing, achieving up to a fourfold increase in spatial density. The reconstruction approach estimates local wavefront characteristics at positions between physical electrodes using electrogram morphology and weighted contributions from surrounding signals within a narrow temporal window. Spatial interpolation combined with model-based regularization enabled recovery of intermediate activation points and more detailed delineation of propagating wavefronts. Reconstruction performance was evaluated by comparing the estimated heterogeneity maps to the simulated ground truth, yielding sub-millimeter spatial deviations. The sensitivity of the reconstruction accuracy to inter-electrode spacing was further tested using different array configurations, with lower density, comparable to clinical mapping catheters.

Our findings show that reconstruction of conduction heterogeneity maps at high-resolution from standard-resolution recordings can enhance the spatial characterization of atrial tissue, offering personalized insights into the mechanisms that drive arrhythmia.