Research On Cardiac Motion Estimation Based On Echocardiographic Speckle Tracking

Cardiac disease has become a major threat to the safety of human life in the 21st century.The number of patients suffering from cardiovascular disease in China has reached 330 million.It is particularly important to timely detect and intervene in the early stage of disease development.Ultrasonic imaging is one of the most effective techniques for monitoring cardiac health.It can give doctors a visual view of a patient’s heart tissue and movement.Ultrasound image is composed of fine band speckle reflected by cardiac tissue.The uniqueness and stability of speckle make the movement state of speckle synchronize with the real movement state of cardiac tissue.Therefore,the study of echocardiographic speckle can provide support for clinical medicine decision making,which has important theoretical and practical significance.In this paper,based on the mainstream research path of echocardiography,a method of automatic segmentation of left ventricle in ultrasonic image is proposed.Secondly,an improved CLG(combined local-globa)optical flow algorithm based on filtering is proposed to track the speckle in echocardiography.Finally,the movement estimation of heart is carried out according to the tracking results.The main work of this paper is as follows:1)The physical basis and background of echocardiography are introduced in detail.On this basis,the formation principle of echocardiography speckle is explained,and the feasibility of speckle tracking technology is analyzed.An automatic segmentation method of left ventricle based on Canny edge detection was proposed.The edge detection technology was used to determine the boundary of the ventricular tissue in the image,and the initial contour of the contour iteration was determined by combining the location model of the junction point,so as to avoid the error caused by different manual labeling of the initial contour.The Snake active contour model was used to iterate the initial contour,and finally the segmentation of the left ventricle was achieved.2)An improved CLG optical flow algorithm based on filter is proposed.Based on the global and local CLG method,the median filter is used to change the anisotropic diffusion coefficient,and the data item and the smoothing term are improved with the combination of the bilateral filter.The vector optical flow field is given by the fixed point iteration method.Based on the display of image tracking points and the numerical analysis of AEE and AAE errors,the detection results of the improved optical flow algorithm are compared qualitatively and quantitatively with the truth values of public data sets through experiments,which shows the effectiveness of the proposed method in precision.3)The mechanical parameters of myocardium were analyzed by echocardiography speckle tracking results.At the same time,the data of the myocardial strain in the data set were calculated to show the change of the left ventricular movement in the form of a chart.Finally,it was compared with the true point strain curve to further prove the reliability of the speckle tracking technology.