Image reconstruction for Magnetic Induction Tomography and Preliminary simulations on a simple head model.

Magnetic Induction Tomography (MIT) of biological tissue is a contactless and noninvasive method to image the complex conductivity distribution or its changes inside a human body. In this paper, 2D (two-dimensional) simulations of the forward problem in MIT are performed by finite element method and an inverse solver based on regularized Newton-Raphson method is developed to reconstruct MIT images from the simulated measuring data. Preliminary MIT simulations on a simple head model are presented. Images can be reconstructed successfully with regularized Newton-Raphson method using the simulated measuring data. The preliminary simulation results show that a better image quality could be obtained by the increase of the number of the coils or the exciting frequencies. And low-conducting skull will exert little influence on the MIT of the brain. Future image reconstruction from real measuring data of physical models is expected and the preliminary simulation results on the brain prove MIT a hopeful and advantageous method on imaging the impedance of the human brain.