Prior dielectric spectroscopy studies have demonstrated that an order of magnitude contrast exists in the dielectric properties of normal and malignant breast tissue at frequencies up to approximately 1 GHz. This contrast serves as the physical basis for confocal microwave imaging, a promising new screening modality that recently has been proposed for early-stage breast cancer detection. In this paper, we present robust signal processing techniques for creating microwave images of the internal breast tissue. To test these image reconstruction algorithms, we have conducted detailed finite-difference time-domain simulations of the confocal microwave imaging sensor applied to the breast. Reconstructed images of the breast models illustrate the potential of confocal microwave imaging for detecting small malignant tumors in the breast. Our current work involving open-ended coaxial probe measurements on freshly excised breast biopsy tissue specimens extends our knowledge of the dielectric properties of breast tissue into the microwave regime and includes a characterization of normal, malignant, and benign breast tissue. This data provides important insights that will enable the development of enhanced detection and discrimination algorithms to be used in confocal microwave imaging.
展开▼
