超声血流的无创伤检测和医学信号的特征提取

人体信号是随机性和背景噪声都很强的复杂信号。文章首先研究了利用超声Doppler技术定量检测血流速度的方法，然后将一些现代信息处理中的新理论、新方法引入医学超声的信息处理，为医学超声信息的特征提取提供了新的手段。这些理论包括：分形、数学形态学、数量化、小波变换、极点轨迹和血管传输线模型等。文章最后阐述了利用上述新方法作为技术核心而研制的三套应用系统；肺动脉血液动力学参数的无损估测系统、彩色编码的声谱系统和超声血流定量检测系统。

Ultrasound Blood Flow Non invasive Detection and Characteristic Extraction of the Medical Signals

Signals from the human body are stochastic, complicated and with strong environmental noise. In this paper, the quantitative blood flow velocity detection using the Doppler ultrasound technique was firstly presented. Then several new theories and approaches of the modern information processing were introduced into the medical ultrasound field, which provided new means to extract the characteristic of the medical ultrasound information. These theories include the fractal, the morphology, the quantification theory, the wavelet transform, the pole track, the transmission line of the vessel, and so on. Finally, three application systems were designed using these new methods as the technology cores: the non invasive detection system for hemodynamic parameters of the pulmonary circulation, the color encoded spectrogram analyzer and the ultrasound quantitative detection system for the blood flow velocity.