EM techniques for the detection of breast cancer

Radar based microwave imaging technique has great potential for breast cancer detection. It has been demonstrated by recent experiments that the data adaptive radar microwave imaging reveals good image resolution. However, purely scattered energy based image reconstruction methods can face major challenges in differentiation since the fibroconnective and glandular tissues which could have similar dielectric properties as that of malignant tissues. It has been proposed that the morphological features of the tissue can differentiate the benign from malignant tissues due to their irregular, spiculated shapes. Thus to efficiently differentiate the healthy, benign and breast tissues, this paper proposes a method that combines a novel data-independent beamformer called MWDAS method with Matrix Pencil (MP) method to first localize the suspicious region and then extract the resonant frequency and damping factor of complex natural resonance (CNR) of the malignant tissues. We employ FD-TD based computational 2D and 3D breast models for testing our method.