Time-of-arrival calibration for improving the microwave breast cancer imaging

Abstract

In radar based confocal microwave imaging for breast cancer detection, recorded data are synthetically focused to a confocal point within the breast. This is the basis for both data-independent and data adaptive methods to form the breast image and can be enhanced by multistatic approach. This approach inherently assumes that the propagation velocity depends only on the average dielectric property of the breast. However, in real cases, the breast tissues are inhomogeneous and therefore the propagation velocities vary for different propagation paths. Thus, use of an average propagation velocity can result in false localization. This paper proposes an auto-calibration method to compensate the time-of-arrival from confocal point to the receiving antennas. We demonstrate using simulations on FDTD numerical breast phantoms that the proposed method helps to form an enhanced image for inhomogeneous breast.