Signal Processing And Image Discrimination Of Breast Cancer Detection Data Under Near-Field Microwave Sounding

Breast tumor is one of the most common cancers of women nowadays, and does serious harm to health. As the incidence of breast cancer rises continuously, early diagnosis of breast tumor becomes more and more important. It will greatly enhance the success of therapy. Therefore, early breast tumor detection technology has been an international highlight of medical imaging science. Traditional X-ray mammography and ultrasonic have shown some deficiencies in sensitivity, security and comfortableness. As a new measurement technique, microwave near-field detection has attracted a lot of attention for providing sufficient functional information.Microwave near-field breast tumor detection implemented by transmitting microwave pulse to the breast tissue and acquiring target back wave. The imaging is obtained by analyzing the characteristics of back wave. Microwave Near-Field Breast Tumor Detection relies on differences in electrical properties between malignant and fatty tissues. The technology has such following excellent features:low radiation energy of electromagnetic signal, a large amount of objective carrier information, and high resolution and Millimeter level accurate location.This thesis is a part of Development Fund of Shanghai Science and Technology key project of basic research——"Breast Tumor Microwave Near-Field Imaging and Early Diagnosis of Breast Cancer (Project Number:08JC1409200).'The characteristics, advantages and disadvantages of common breast cancer methods are reviewed. We propose the method of breast cancer detection under microwave near-field sounding environment. In this paper BRATUMASS's structure (breast tumor sensor system), the principle of detection space partition, back-wave acquisition and signal processing under strong noise background are introduced.Based on the experiment orientation, the principle and algoritlum of inversion imaging, the distribution of sampling points are described in detail.This dissertation gives out the implementation of the imaging platform for breast cancer detection data under near-field microwave sounding. All module functions of the software and operation steps are introduced concretely. Carrying out Hilber-Huang transformation to extract all instantaneous frequency components of signal, and set up value adoption range of inversion calculation, image output is obtain through iteration. User can get important pathological information from the image.Clinical cases imaging show that Breast cancer detection terminal under near-field microwave sounding is helpful to detect lesions and confirm the position of lesions. In the figure, color grading imaging corresponding to the dielectric constant difference between malignant tumor and normal breast issue. By analysis signs of figure, tumor place and tumor type can be pre-judged. As a medical preliminary judgment, Breast cancer detection terminal under near-field microwave sounding is a useful diagnosis reference.