Zhang Yufeng, Gao Yali, Wang Le, Chen Jianhua, Shi Xinling
Department of Electronic Engineering, Information School, Yunnan University, Kunming City, Yunnan Province 650091, China.
IEEE Trans Biomed Eng. 2007 Sep;54(9):1631-42. doi: 10.1109/TBME.2007.891936.
Doppler ultrasound systems, used for the noninvasive detection of the vascular diseases, normally employ a high-pass filter (HPF) to remove the large, low-frequency components from the vessel wall from the blood flow signal. Unfortunately, the filter also removes the low-frequency Doppler signals arising from slow-moving blood. In this paper, we propose to use a novel technique, called the empirical mode decomposition (EMD), to remove the wall components from the mixed signals. The EMD is firstly to decompose a signal into a finite and usually small number of individual components named intrinsic mode functions (IMFs). Then a strategy based on the ratios between two adjacent values of the wall-to-blood signal ratio (WBSR) has been developed to automatically identify and remove the relevant IMFs that contribute to the wall components. This method is applied to process the simulated and clinical Doppler ultrasound signals. Compared with the results based on the traditional high-pass filter, the new approach obtains improved performance for wall components removal from the mixed signals effectively and objectively, and provides us with more accurate low blood flow.
用于血管疾病无创检测的多普勒超声系统通常采用高通滤波器（HPF），从血流信号中去除来自血管壁的大的低频成分。不幸的是，该滤波器也会去除由缓慢流动的血液产生的低频多普勒信号。在本文中，我们提出使用一种名为经验模态分解（EMD）的新技术，从混合信号中去除壁成分。EMD首先将一个信号分解为有限且通常数量较少的名为固有模态函数（IMF）的单个成分。然后，基于壁血信号比（WBSR）的两个相邻值之间的比率，开发了一种策略，以自动识别和去除对壁成分有贡献的相关IMF。该方法应用于处理模拟和临床多普勒超声信号。与基于传统高通滤波器的结果相比，新方法在从混合信号中有效且客观地去除壁成分方面取得了改进的性能，并为我们提供了更准确的低血流信息。
