A method for solving the three‐dimensional inverse problem in electrocardiography using the finite‐element method

Abstract

AbstractThe solution of the inverse problem in electrocardiography, if it can be obtained, is clinically significant since information on cardiac excitation can be obtained non‐invasively from the surface electrocardiogram. This paper describes a solution for the three‐dimensional inverse problem in electrocardiography and examines the possibility of clinical application based on a numerical example for the human torso model. In the inverse problem in electrocardiography the potential transfer matrix relating the epicardial potential distribution to the body surface potential distribution is determined, which enables estimation of the former from the latter. Owing to the smoothing effect in the human thorax, however, the potential transfer matrix is modified such that the calculated epicardial potential distribution becomes oscillatory. In our study, this problem is remedied by using the conditional least‐squares filter and applying Tikhonov's regularization. The relation between the number of electrodes, measurement accuracy of the body surface electrocardiograph and the error in estimation of the epicardial potential distribution is discussed. Based on this discussion, it is concluded that estimation of the epicardial potential distribution with practicable accuracy requires at least 180 electrodes and three‐place measurement accuracy for the body surface electrocardiogram.