Spatiotemporal analysis of cardiac electrical activity

Electric measurements from the heart are a valuable source of information in both normal and pathological electrophysiology. They also present special signal processing challenges to the researcher or clinician intent on extracting this information. For example, spatial sampling is often limited and almost always irregularly organized. Sources of cardiac signals are time varying, distributed within a complex, anisotropic excitable medium. The goal of cardiac signal analysis is to describe spatial and temporal features of activation and relaxation of the heart tissue; both spatial and temporal views have importance and in many cases merging of the two is necessary to reveal the relevant information. In this presentation, we cover some of the techniques that electrophysiologists have developed to analyze and interpret cardiac electric potentials. Specific examples include interpolation and statistical estimation, time and space based methods for locating the cardiac wavefront, and the use of explicit solutions of boundary value problems to predict cardiac sources from ECG measurements. We also outline some of the challenges remaining in cardiac signal processing.