Parametric Study of Time-Reversal Imaging for the Detection of Tumors in MRI-derived Breast Phantoms

In this paper, the performances of two time-reversal (TR) based algorithms namely time-reversal multiple signal classification (TR-MUSIC) and decomposition of time reversal operator (DORT under its French acronym) for detecting breast cancer through UWB microwave imaging, are assessed. These assessments are carried out by changing the location of a hypothetical spherical tumor in an MRI-based numerical breast phantom through simulations in CST microwave studio environment. The recorded data is processed by the two TR-based procedures which result in the final images of cancerous breast tissues. In all cases, data acquisition has been conducted in 2 steps with and without a tumor inside the breast. During each step, UWB pulses within the wideband frequency range of 2-5 GHz are transmitted by 16-element antenna array which also receives backscattered signals from the breast tissue. In this work, we show that the implemented method can locate tumors with high precision, whether the tumor is located in the middle of the breast phantom, near to the skin or even close to the muscle wall which possesses a high reflectivity.