Microwave inverse scattering for breast tumor detection using finite difference frequency domain modeling

Abstract

We have implemented a new algorithm for the electromagnetic inverse scattering problem in inhomogeneous media using a MATLAB-based finite difference frequency domain (FDFD) forward model. This algorithm, referred to as FDFD-based inverse method, constructs a near-linear expression for inversion from the matrix form of the forward model and uses a vector-based Born approximation without employing an a priori Green's function. This algorithm is easily implemented and is robust to the heterogeneity of the background. 2D microwave breast tumor detection is investigated using an MRI-based breast model for the purpose of checking the application in highly heterogeneous material.