全方向M型心动图运动状态提取新方法

摘要

传统上,在全方向M型心动图上提取心脏运动参数采用直接换算的方法,然而该方法会使测量噪声放大,无法获得高精度的结果. 在之前多尺度Contourlet变换算法提取心动图运动曲线的基础之上,考虑心脏运动与超声运动的图像特点,通过在心动图上测量相邻几个周期的相对位移情况,建立心动图运动模型,最后通过自适应渐消卡尔曼滤波算法,获得比传统算法精度更高的结果,并通过实验,计算各个算法获得结果的平滑程度,证明了论文算法比传统算法具有更高的精度.%The method of Omni-directional M-mode Echocardiography State Extraction is conversion directly on traditional. However, this method made the measurement noise amplification, and could not obtain an exact results. Paper measured several cycles of relative displacements on the map base on the the detection algorithm with continuous Contourlet transform. Paper considered the characteristics of cardiac motion, and modeled the echocardiography. At last, paper extracted motion state base on adaptive fading Kalman filter, and got a more accurate result. In final, paper calculated the smooth curve degree to prove and proof.