Compressed sensing for three-dimensional microwave breast cancer imaging

In this paper, we report on a compressed sensing framework using sparse reconstruction for microwave imaging of human breast cancer. Therefore, we have implemented a simulation environment that enables the investigation of novel image reconstruction concepts under plane wave assumption. Results are shown for orthogonal matching pursuit, and two beamforming methods namely the commonly used delay-and-sum (DAS) method and a beamforming technique based on the coherence factor (CF). Using compressed sensing allows a reduction of transmitter and receiver elements and at the same time leads to an improved image quality. On top of that, we have investigated the robustness of the image reconstruction performance through a statistical analysis using different noise levels.