论文部分内容阅读
有效抑制由静态或慢动组织引起的杂波是诊断超声彩色血流成像技术研究的关键问题之一,不充分的杂波抑制将导致多普勒平均频率估计趋同于杂波成分特性,从而导致潜在的血流信息描述误差。本文在研究广义壁滤波器设计原理基础之上,比较了三种不同类型壁滤波器——投影初始化有限冲击响应滤波器(Prj-IIR)、多项式回归滤波器(Pol-Reg)与基于特征值分解(Eigen-based)滤波器的设计方法差异,并对其杂波抑制性能进行了仿真对比分析。血流平均频率一维自相关估计结果表明,投影初始化IIR壁滤波器与多项式回归壁滤波器杂波抑制性能相近,在平稳杂波环境下可实现血流平均速度的精确估计,当多普勒采样向量包含样本数据增加时,改变设计参数可进一步改善杂波抑制能力,而基于特征值分解的壁滤波器则可对非平稳杂波进行抑制,进一步改善低速血流的估计精度,而且在多普勒矢量包含样本数小于10的情况下,计算复杂度并无显著增加。
Effective suppression of clutter caused by static or slow tissue is one of the key issues in the research of diagnostic ultrasound color flow imaging. Incomplete clutter suppression will lead to the convergence of Doppler average frequency estimation to clutter components, resulting in Potential blood flow information describes the error. Based on the research of generalized wall filter design principle, three different types of wall filters are compared: Prj-IIR, Pol-Reg and filter based on eigenvalue Eigen-based filter design method differences, and its clutter suppression performance simulation comparative analysis. One-dimensional autocorrelation estimation of average blood flow frequency shows that projection initialized IIR wall filter and clutter suppression performance of the polynomial regression wall filter are similar, and accurate estimation of average blood flow velocity can be achieved under stationary clutter. When Doppler However, eigenvalue decomposition based wall filters can suppress non-stationary clutter and further improve the estimation accuracy of low-speed blood flow when the sample vector contains more sample data. Moreover, when the sample data is increased, changing the design parameters can further improve the clutter suppression ability. In the case that the number of samples contained in the vector is less than 10, the computational complexity does not increase significantly.