Beamspace time reversal maximum likelihood estimation for microwave breast imaging

We consider maximum likelihood based beamspace time reversal beamforming for breast cancer localization. We reduce the computational burden of maximum likelihood estimation through reduced dimensional beamspace processing. Beamspace processing also provides additional beamspace gain which contributes to suppress strong clutter effects. We collect multistatic scattering fields through FDTD simulation and further process it in beamspace for maximum likelihood based time reversal imaging. The imaging technique is used to localize a small tumor in a dense breast. It is observed that the proposed imaging technique can localize tumors unambiguously even in dense breast phantom.