Ultra-wideband microwave imaging of breast cancer tumors via Bayesian inverse scattering

We develop a new algorithm for ultra-wideband (UWB) microwave imaging of breast cancer tumors using Bayesian inverse scattering. A key feature of the proposed algorithm is that constitutive properties of breast tissues are reconstructed from scattered UWB microwave signals together with the confidence level of the reconstruction. Having such confidence level enables minimization of both false alarms and missed detections. Results from the application of the proposed algorithm demonstrate the accuracy in estimating both location and permittivity of breast tumors without the need for a priori knowledge of pointwise properties of the background breast tissue.