Estimation of the Frequency-Dependent Average Dielectric Properties of Breast Tissue Using a Time-Domain Inverse Scattering Technique

Abstract

Ultrawideband (UWB) microwave radar imaging techniques for breast cancer detection typically require estimates of the spatially averaged dielectric properties of breast tissue. We propose an algorithm for estimating patient-specific, frequency-dependent average dielectric properties from scattered UWB microwave signals. The algorithm is a variation of an iterative finite-difference time-domain (FDTD) technique for solving the time-domain inverse scattering problem. The assumption is made that the breast consists of two homogeneous regions: skin and underlying tissue. This two-medium geometry results in an inverse scattering problem that is not as ill-posed as those typically encountered in microwave tomography because the number of unknowns being sought is very small. We test the performance of the algorithm on data simulated using anatomically realistic 2-D numerical breast phantoms