An Innovative Approach to Automatically Detect and Interpret Salient Spatiotemporal Features of a Numeric Field: A Case Study in Electrocardiographic Imaging

摘要

The last decade has witnessed major advancements in the direct application of functional imaging techniques to several clinical contexts. Unfortunately, this is not the case of Electrocardiology. As a matter of fact, epicardial maps, which can hit electrical conduction pathologies that routine surface ECG's analysis may miss, can be obtained non invasively from body surface data through mathematical model-based reconstruction methods. But, their interpretation still requires highly specialized skills that belong to few experts. The automated detection of salient patterns in the map, grounded on the existing interpretation rationale, would therefore represent a major contribution towards the clinical use of such valuable tools, whose diagnostic potential is still largely unexploited. We focus on epicardial activation isochronal maps, which convey information about the heart electric function in terms of the depolarization wavefront kinematics. An approach grounded on the integration of a Spatial Aggregation (SA) method with concepts borrowed from Computational Geometry provides a computational framework to extract, from the given activation data, a few basic features that characterize the wavefront propagation, as well as a more specific set of features that identify an important class of heart rhythm pathologies, namely reentry arrhythmias due to block of conduction.