Early time detection of the breast cancer will lead to a more efficient treatment and comfort of the patient. Currently the gold standard method in screening breast cancer is x-ray mammography which employs a low dose of ionizing radiation to create a high resolution image of the compressed breast. Although this method has been shown to be able to save lives, it still has shortcomings, the most important among which is the use of ionizing radiation that prevents this method to be used as a regular screening method. Hence, other modalities for the possible use in breast cancer detection has been investigated. Among them, UWB microwave pulse has been shown to be promising in detecting breast tumor while it has no known health risk for the patient. But a major problem in applying this method for breast tumor detection is the strong backscatter produced by the skin surface which can mask the weak tumor signature reflected from inside the breast medium. Several methods are proposed to solve this problem but none can effectively remove this artifact in general plus the deterioration they introduce to the tumor response. In this study a new method exploiting frequency domain information to overcome this problem is proposed. In this method the frequency response is analyzed and skin related information is removed and the time domain skin-removed signal is obtained from frequency response. Applying this method to simulated models of the breast shows that it can effectively remove the skin backscatter without introducing deterioration to the tumor response. Also the effectiveness of this method in removing the skin reflection is compared with the current methods. Results show that this algorithm can be considered as a promising method for skin artifact removal. However, more studies are needed to test the capability of this method in more realistic cases.

75 p.