Data Fusion for Inverse Electrocardiography: Synthesis of Signals from Multiple Sensor Types and Locations

Rose-Hulman Institute Of Technology, Terre Haute IN

Investigators

Abstract

0001315 Throne Knowledge of how electrical potential patterns on the heart surfaces evolve over time can be an extremely valuable tool in the diagnosis and treatment of a wide variety of cardiovascular diseases. Hence, researchers have proposed numerous minimally invasive approaches for inferring cardiac electrical activity from measurements of electrical activity made at more accessible locations. In these types of inverse problems the goal is to estimate an input (the surface electrical potentials) based on a measurement of the output (the electrical activity measured at a less invasive location). These particular inverse problems are extremely ill-conditioned, and consequently small errors in the measurements cause enormous (hundreds of times larger) errors in the estimated cardiac potentials. The overall goal in this project is to employ data fusion-- systematic combination of signals from multiple sensors, types of sensors, and sensor locations-- to improve the potential estimates on the interior and exterior surfaces of the heart. The development of minimally invasive, stable inverse techniques will substantially affect the treatment of cardiac disease. In addition, the data fusion techniques developed for these inverse problems will impact research in such diverse fields as identifying sources of electrical activity within the brain, structural damage identification, electronics cooling, materials characterization, and online machine tool wear monitoring.