Discussion of Current Distribution of Exercise 3D MCG Using an Inhomogeneous Human Volume Conductor

Abstract

In this paper we present the analysis of 3D measurements of the magnetocardiogram by means of current source reconstruction. The data considered are rest and exercise stress data of test subjects. The source model used to describe the heart activation is a distributed source model, which allows to follow how the heart activation spreads over the myocardium. We developed a four compartments human torso model including the heart cavity and two lungs. Different values of the compartment conductivities are used to study the effect of the inhomogeneities on the reconstructed sources. The comparison between rest and stress data is done by means of correlations coefficients that indicate how close the reconstructed sources are in terms of dipoles moments, since in the source model considered the dipoles positions are fixed and depend on the source space model geometry. The same analyses are repeated using the 3D magnetic data and only the normal component, to investigate the additional information given by the tangential components. The results show that the torso and heart conductivities have a stronger influence than the one of the lungs, and that the 3D data are less sensitive to the changes of the compartment conductivities.

Discussion of Source Reconstruction Models Using 3D MCG Data

“…The model used for the CSR calculations is the one showed in fig. 1, where each triangulated surface has 586 triangles (16) , and the potential is supposed to vary linearly over each triangle.…”

“…To compare single dipole and distributed source localizations, an equivalent dipole can be calculated for the distributed sources (16) , having as position the center of mass of all dipoles and as amplitudes the average of all amplitudes: where J(t) is the average of all amplitudes, P(t) is the center of mass, i is the dipole index, N D is the total number of dipoles, |J i (t)| is the amplitude of the i-th dipole at the time t and P i its position. Whatever the number of dipoles N D is, using the equations (4) and (5), a single equivalent dipole can be found from a cluster of dipoles.…”