Abstract

In the last years, Electrocardiographic Imaging (ECGI) systems have been developed to support the study of cardiac arrhythmia. In order to clinically validate those systems, it is often necessary to use the estimated epicardial potentials, which correspond to the electrode-recorded unipolar electrograms (EGMs) in electrophysiological studies, with the features obtained from bipolar EGMs that are often used to establish the clinical criteria running in the daily practice. In this work, the properties expected in the bipolar EGMs were scrutinized, in terms of the amplitude of the neighbor EGMs and the spatial consistency of closely recorded EGMs for each given EGM in a cardiac location. Several possible Digital Signal Processing Operators (DSPOs) were explored as possible basis for the calculation of bipolar EGMs from the unipolar EGMs from ECGI, namely, maximum neighbor amplitude, minimum neighbor distance, and random neighbor. For each of these DSPO, the impact on the amplitude for each estimated bipolar EGM were analyzed. Our results show that the optimal among these DSPO to calculate the bipolar EGMs associated to a given spatial location in the cardiac epicardium is obtained when using the neighbor with maximum EGM amplitude, as supported by its usefulness to identify scarred regions and by spatial consistency properties. These results allow to establish a comparison of ECGI estimated potentials with the usual electrophysiological criteria when these are established in terms of bipolar EGMs.