Three-Dimensional Linear Sampling Applied to Microwave Breast Imaging

Microwave imaging is a promising technique in biomedical imaging. Full-wave quantitative inversion has much potential but is computationally very demanding due to the non-linearity of the problem while qualitative approaches mostly use approximations that do not comply with contrasted biomedical scatterers. In contrast, the linear sampling method is capable of detecting anomalies in a known background by solving only a linear problem. Although only a qualitative image is obtained in the form of an indicator function that assumes low values inside and large values outside the anomaly, it may be suitable for use in breast cancer imaging. This paper reports a preliminary study of the feasibility of linear sampling for this application. 1