A new regularization approach to reconstruct tissue properties using a priori information

A new microwave imaging method for early detection of breast cancer is presented in this paper. Solving the electromagnetic inverse scatter problem involves estimating the distribution of dielectric properties, which is both nonlinear and ill-posed. In order to get an acceptable result, we make use of some a priori information in regularization, which is obtained from the currently effective imaging system (e.g. CT and MRI ) and the recent electrical properties study of biological tissues: we assume knowledge of the boundary and the dielectric properties of the various tissues to optimize the result. The inverse scattering problem formulated as a nonlinear optimization problem is numerically solved using Distorted Born Iterative Method (DBIM). We show that the new approach can produce a more accurate estimate of the distribution of dielectric properties, and reduce the iterative time.