Abstract

Several aspects of the forward and inverse problems of electrocardiography are investigated through the use of digital computer models. Two forms of a fixed location, variable moment, 20-dipole cardiac model of QRS are developed from actual cardiac excitation data. One form uses time-varying orientation dipoles; the other uses fixed orientation dipoles. An electric multipole expansion (EME) cardiac model employing the dipole, quadrupole, and octupole terms is also developed and used as an equivalent forward and inverse cardiac model. Two realistically shaped torso models are used. The homogeneous torso has uniform conductivity; the inhomogeneous torso contains realistically shaped lung regions with reduced conductivity. It is found that when the EME model is used as an equivalent forward cardiac model, it can accurately represent the actual 20-dipole cardiac model in the homogeneous torso. Limb leads are accurately represented by the dipole terms alone while the precordial leads require the quadrupole and octupole terms. It is also found that while the lung regions have little effect on the ECG's produced by the models, these regions can have a significant effect on the inverse solutions for certain dipoles in the 20-dipole cardiac model. These lung regions appear to have a much smaller effect on the dipole terms in the EME model. Solutions of the inverse problem for the terms in the EME model indicate that when a limited number of measurements are used, the best results can be obtained by uniform distribution of the measurements over the torso.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call