Reconstruction of the anisotropic complex conductivity distribution of a human thigh in three dimensions

In electric impedance tomography (EIT), a set of electrodes measures surface potentials resulting from known currents injected into the target region. Using a correct model of the target region, these potentials can be simulated very accurately, too. Since injected currents spread out in three dimensions, simulations in three dimensions are inevitable. In addition, the anisotropic nature of electric material properties of biological tissue should not be neglected to obtain appropriate results. The aim of this work is to reconstruct the anisotropic complex conductivity distribution of a human thigh. "Pseudomeasurements" making up for EIT data were available. A finite element model of a human thigh based on magnetic resonance images was generated. A classical Gauss-Newton algorithm with regularization was used to match the real material values and the ones fed into the simulation. Results for a simple, but unrealistic isotropic case as well as for a complicated anisotropic case are presented and discussed