Three algorithms and three electrode models for electrical impedance image reconstruction

Electrical impedance tomography is a technique for reconstructing the spatial distribution of the parameters characterizing the electrical properties of tissues within a body. These parameters, conductivity and relative permittivity, cannot be obtained by other imaging modalities. In this paper, the simulated annealing, the genetic algorithms, and the Newton-Raphson method are developed for electrical impedance image reconstruction. From the results, both the simulated annealing and genetic algorithm based methods are demonstrated to be feasible for producing the dynamic image of resistivity distribution. In order to reduce the number of degree of freedom in these two methods, a guided technique based upon the voltage measurements is also designed. In addition, a modified Newton-Raphson method is also proposed for improving the image contrast. Three electrode models for describing the effect of electrode are also investigated. Afterwards, A simple gap electrode model for its best performance is also included in the FEM forward solver.