Computer models of electric fields in the human torso have numerous applications and are useful as a substitute for invasive measurements [1, 2, 3]. A few examples of modeling applications include electrical impedance tomography , the study of bioelectric fields from the heart [5, 6], and design and optimization of implantable defibrillators . As the cost of high-speed processing and computer memory has decreased, it has become increasingly practical to model larger and more realistic problems. In each modeling scenario encountered, however, the details of the physiology are still too numerous and too complicated to be fully modeled, even with the most powerful computational systems currently available. In the specific case of the electric fields arising in the thorax due to the heart, the level of detail of the two most important aspects of the simulation--the geometrical model of the thorax and the conductivity values assigned to each region of the geometry--is still very limited. The simplest assumption is that conductivity is homogeneous (the same material properties apply throughout the volume) and isotropic (material properties do not depend on the orientation of field vectors). More detailed models attempt to characterize tissue inhomogeneity and anisotropy by including one or more inhomogeneous regions. However, while the role and relative importance of inhomogeneities for specific electrocardiographic situations has been the topic of many different thorax model studies [8, 9, 10, 11, 12, 13, 14, 15]; there is currently no consensus on the subject among researchers. The goal of the study reported here was to evaluate, through the use of advanced numerical techniques and a detailed model of the human torso, the role of inhomogeneities (including anisotropic ones) in computing forward solutions from epicardial heart potentials.
Several specific aspects of electric fields in the thorax further complicate the inclusion of inhomogeneities in model geometries. The first is the fact that torso conductivities cover a wide range: the electrical conductivity of blood, for example, is over 47 times larger than that of bone . In addition, the inhomogeneity of the electric field within the thorax, together with the asymmetry of its geometric structures, suggest that not only the values but also the locations of conductive inhomogeneities will influence field calculations. Finally, the field distribution within the thorax during a heart beat is dynamic in nature. While the frequencies that arise are considered low enough to treat the problem as quasi-static in nature , the interaction between electric field and resistive tissues will vary in time due purely to spatial variation in the field. Hence, there exists a need to consider the influence of a wide variety of tissues on model calculations and to evaluate them under a dynamic range of conditions.
In this study, we evaluated the effects of selected inhomogeneities and anisotropies on computed body surface potential maps (BSPMs) in a large-scale model of the human thorax. The results can serve to guide the selection of inhomogeneities to include in any geometric model used to compute bioelectric fields. It should be noted that it is currently not possible to measure epicardial and body surface potential distributions simultaneously in humans and thus data for absolute validation are not available. What these results can establish, however, is the relative effect of selectively including or excluding a wide variety of inhomogeneities in a detailed model of the human thorax.