Simultaneous recovery of three-dimensional myocardial conductivity and electrophysiological dynamics: A nonlinear system approach

Despite its recognized essentiality for assessment of patient-specific physiopathological states, passive cardiac tissue conductivity remains largely unexplored in inverse electrocardiography (IECG) studies. In this paper, we present a novel framework for simultaneous recovery of the volumetric myocardial conductivity and electrophysiological dynamics from a posteriori image-derived heart- torso geometry and body surface potential (BSP) measurements that are constrained by a priori electrophysiological system models with uncertain parameters. The main innovation of the effort lies in the introduction of a nonlinear dynamic system estimation paradigm into the IECG problems, which leads to forward modeling of the cardiac electrophysiological processes and inverse recovery based on joint parameter and state estimation of nonlinear dynamic system via unscented Kalman filtering (UKF). Experimental results show that the framework produces accurate and robust three-dimensional (3D) transmembrane potential (TMP) map sequences and conductivity map.