Non-invasive localization of atrial ectopic beats by using simulated body surface P-wave integral maps

摘要: Non-invasive localization of continuous atrial ectopic beats remains a cornerstone for the treatment arrhythmias. The lack accurate tools to guide electrophysiologists leads an increase in recurrence rate ablation procedures. Existing approaches are based on analysis P-waves main characteristics and forward body surface potential maps (BSPMs) or inverse estimation electric activity heart from those BSPMs. These methods have not provided efficient systematic tool localize triggers. In this work, we propose use machine learning techniques spatially cluster classify foci into clearly differentiated regions by using P-wave integral map (BSPiM) as biomarker. Our simulated results show that with similar BSPiM naturally non-intersected new patterns could be correctly classified accuracy 97% when considering 2 clusters 96% 4 clusters. also suggest number is feasible at cost decreasing accuracy.