超宽带微波检测早期乳腺癌信号处理研究

发布时间：2018-02-04 22:43

本文关键词： 超宽带微波早期乳腺肿瘤频域波束形成肿瘤特征信号提取集成经验模式分解　出处：《天津大学》2014年硕士论文　论文类型：学位论文

【摘要】：近年来,乳腺肿瘤越来越成为威胁全世界女性的生命健康的一种疾病,其造成的患者的死亡率一直处于所有癌症的第一位。因而乳腺癌的治疗与检测引起了全世界的关注,全世界各个国家均投入了很大的精力与物力进行乳腺癌的治疗与检测的研究。最终临床医学研究证实,在乳腺肿瘤早期检测出肿瘤同时进行治疗是有助于提高乳腺肿瘤患者的成活率的。因而目前有很多的检测乳腺肿瘤的方法与手段,但是他们都有一定的缺点,使得乳腺癌早期检测的实施有了一定的障碍。然而,超宽带微波检测早期乳腺肿瘤的方法具有价格便宜,检测过程无痛苦,对患者身体无辐射等很多的优点,因而受到全世界的广泛关注。因而,在这个背景下,本文从介绍超宽带微波检测入手,介绍了真实乳房的结构以及乳房模型的导出与建立,主要介绍了超宽带微波检测早期乳腺肿瘤系统中所使用的三种模型,有二维平面模型,三维长方体模型,以及基于核磁共振图像导出的二维乳房模型。其中使用核磁共振图像导出的乳腺肿瘤模型是最为接近真实乳房的一种模型。研究了超宽带微波频域波束形成成像的具体过程以及结果,同时介绍了两种不同的肿瘤特征信号的方法。详细介绍了超宽带微波频域波束形成成像的基本原理,基于二维平面模型进行的频域波束形成算法的具体实施过程以及成像结果。本文主要研究了两种提取肿瘤信号的方法,一种是先提取基准信号,然后通过减去基准信号得到肿瘤特征信号,这种方法是基于三维长方体模型的实验,是采用天线阵列设计的方法进行的肿瘤基准信号提取,进而提取肿瘤特征信号。另一种肿瘤特征信号提取的方法是直接从得到的信号中提取肿瘤特征信号。这种方法是基于集成经验模式分解的时频分析方法对得到的信号进行分解处理,从而得到有用的肿瘤特征信号,并进行成像。
[Abstract]:In recent years, breast cancer has become a disease that threatens the life and health of women all over the world. The death rate of patients caused by it has been the first of all cancers. Therefore, the treatment and detection of breast cancer has attracted worldwide attention. Countries all over the world have invested a lot of energy and material resources to study the treatment and detection of breast cancer. It is helpful to improve the survival rate of breast tumor patients by simultaneous treatment in the early detection of breast tumor. Therefore, there are many methods and means to detect breast tumor at present, but they all have some shortcomings. The implementation of early detection of breast cancer has some obstacles. However, ultra-wideband microwave detection of early breast cancer has many advantages, such as low price, painless detection process, no radiation to patients and so on. Therefore, under this background, this paper begins with the introduction of UWB microwave detection, introduces the structure of real breast and the derivation and establishment of breast model. This paper mainly introduces three kinds of models used in the detection of early breast tumors by UWB microwave, including two dimensional plane model and three dimensional cuboid model. The breast tumor model derived from MRI image is the closest to the real breast model. The ultra-wideband microwave frequency domain beamforming imaging is studied. And the results. At the same time, two different methods of tumor characteristic signal are introduced, and the basic principle of ultra-wideband microwave frequency domain beamforming imaging is introduced in detail. The implementation process and imaging results of the frequency domain beamforming algorithm based on two-dimensional plane model. This paper mainly studies two methods to extract tumor signal, one is to extract the reference signal first. Then the tumor characteristic signal is obtained by subtracting the reference signal. This method is based on the experiment of three-dimensional cuboid model and is extracted by using antenna array design method. Another method is to extract tumor feature signal directly from the obtained signal. This method is based on the time-frequency analysis method integrated with empirical mode decomposition. The signal is decomposed. Thus the useful tumor characteristic signal is obtained and the imaging is carried out.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R737.9;TN911.7

【参考文献】