Effects of electrical impedance measuring methods on two-dimensional tomogram recovery of biological tissues

Abstract

The purpose of electrical impedance tomography is to obtain the electrical impedance distribution in the domain of interest by injecting the currents or applying voltages and measuring voltages or currents via a number of electrodes that are mounted on the boundary of the domain. We investigated the influence of various alternating current injection methods on conductivity allocation recovery in biological tissues. We used 16 electrodes allocated uniformly on a circle perimeter. The research technique includes the mathematical modeling by finite element method with 576 nodes. The current injection was performed through two electrodes located nearby (dipole assignment), opposite (polar assignment) or with a shift by 3 electrodes (a quarter of circle). We registered the potential differences between other electrodes for calculation of the internal conductivity allocation by the finite element method. The study revealed that dipole current injection impoved the sensitivity of the method, and polar injection refined the resolution capability. We used the absolute and difference calculation methods implemented in the programming package of potentials allocation and image reconstruction EIDORS (Electrical Impedance and Diffuse Optical Tomography Reconstruction Software). EIDORS is an open source software system for image reconstruction in the electrical impedance tomography and diffuse optical tomography, designed to facilitate collaboration, testing and new research in these fields. Several numerical examples with inclusion of various convex and non-convex smooth shapes (e.g. circular, elliptic, square-shaped) and sizes are presented and thoroughly investigated. The experiments revealed phantoms at round form discontinuities of conductivity. As an accuracy criterion, we selected mean-square and maximum deviation values of the reconstructed image from the true conductivity allocation. The study showed the advantages, lacks and application fields of dipole, polar and other methods of the current injection. The experiments demonstrated the optimal parameters for reconstruction of internal conductivities at various methods of stimulation. The model with polar electrodes showed the best results by the criterion of maximum deviation. The model with electrodes shifted on a circle quarter revealed the best results by mean-square error criterion.