Quantification of local heterogeneity from activation maps of omnipolar multielectrode recordings

Abstract

Abstract Background and Purpose The study of cardiac arrhythmias requires understanding the heart's electrical conduction patterns. Using multielectrode catheters and other mapping systems have allowed a futher understanding of the wavefront. In particular, novel high-density (HD) grid catheters, enable the recording of omnipolar electrograms, which are purportedly not influenced by catheter orientation, allowing real-time determination of signal propagation direction. We propose a novel heterogeneity metric to quantify the degree of disorganisation in the vector maps from reconstructed omnipolar electrograms, which could potentially be utilised to detect and assess the degree of fibrosis in cardiac tissue. Methods We utilised basal and stimulated recordings from 5 Langendorff-perfused rabbit hearts, where signals were obtained using a multielectrode mapping catheter with 128 electrodes spaced at 1mm. Three recordings were taken for each series, basal at 37°C, plus stimulation performed using a bipolar electrode at 4 Hz and 6 Hz . An angle propagation matrix is computed for the local area of recording, by means of the orthogonal bipolar loops of each clique. A local heterogeneity metric is derived from the summation of angular variations of neighbouring elements. An illustration of the heterogeneity map, and heterogeneity metric for the recording area (H) is illustrated in figure 1. A statistical analysis is performed then to assess the distribution of heterogeneity values. Results Heterogeneity values for each experimental heart, and the distributions of H values obtained according to the stimulation type are shown in figure 2A. The ROC curve obtained from the values of H has an AUC of 0.996, with an optimal threshold value of 0.113, resulting in a confusion matrix presented in 12 TP, 30 TN, 3 FP and 0 FN. The study shows that the proposed heterogeneity metric can quantify the degree of disorganization in the vector maps and can be used to differentiate between basal and stimulated groups with p<0.001. Conclusion This study proposes a novel heterogeneity metric to quantify the degree of disorganization in the vector maps from reconstructed omnipolar electrograms. The methodology was assessed on basal and stimulated epicardial rabbit recordings and found to be effective in differentiating between the two groups with high accuracy. Future studies should focus on the efficiency of assessing fibrotic tissue using the proposed metric.Figure 1Figure 2