Boundary Extraction Based Imaging Method by Incorporating FDTD Based Wavefront Analysis for Microwave Mammography

摘要

Microwave UWB (Ultra Wideband) radar is promising as a non-invasive medical imaging, such as breast cancer or brain stroke detection due to its non-ionizing feature and deeper penetration ability in low loss medium. While various imaging algorithms has been developed, most of them are based on delay-and-sum (DAS) approach, which could not accurately reconstruct the boundary of dielectric object, even if a complete propagation model is given. This paper proposes a new scheme for accurate boundary extraction based on the range points migration (RPM) algorithm, where the candidate wavefront curve is accurately extracted by the FDTD based data generation. The results in numerical simulation assuming highly heterogeneous breast medium case, demonstrate that our algorithm achieves accurate boundary extraction for cancer shape estimation.