Evaluation Of Myocardial Wall Segments By Echocardiography Based On Optical Flow

Ultrasound technology is widely used in clinic because of its advantages of non-invasive,real-time and low-cost.Echocardiogram is often used to evaluate the structure and function of the heart.In echocardiography,the movement of the myocardial wall is one of the most commonly used examination methods for diagnosing cardiac dysfunction in clinic,and also one of the important indicators for evaluating cardiac function.However,due to the limitations of ultrasound physical imaging,the change of patient’s position during the acquisition process and other factors,there are image quality problems such as noise and speckle in echocardiography,which has a certain impact on motion estimation.In addition,as a diagnostic technique of cardiovascular diseases,echocardiography depends largely on the subjective judgment of doctors.With the development of computer vision technology,quantitative analysis of myocardial wall motion in echocardiography with the help of motion estimation can effectively reduce the potential risk.In view of the above problems,the main work of this paper is as follows:(1)Aiming at the problem of blurred boundary in echocardiographic imaging,wavelet analysis is proposed to extract the motion boundary information,which is used as an adaptive smoothing coefficient after function mapping,and added to the optical flow regularization process for iterative calculation,so that the optical flow calculation results are more accurate and the boundary information is more abundant,which can lay a foundation for further quantitative evaluation of myocardial function and auxiliary clinical diagnosis and treatment.(2)Aiming at the problems of image quality such as speckle,noise and shadow in echocardiography and the large amplitude of ventricular wall motion during myocardial contraction,an optical flow calculation method combining multi-resolution analysis and the confidence optimization strategy of optical flow is proposed.The approximate coefficient of ultrasound image is extracted by wavelet decomposition and reconstruction as the confidence optimization coefficient,and as the weight coefficient in the iteration of optical flow pyramid,The objective is to reduce the iterative calculation weight where there is less information,and increase the calculation weight where there is rich texture information.By selecting appropriate pyramid resolution,the accuracy of optical flow calculation under large displacement can be effectively improved.(3)In order to better assist clinical diagnosis,a method of motion estimation of each segment of myocardial wall based on multi-resolution confidence optimization optical flow is proposed.First,the myocardial wall is manually segmented,and then the optical flow method is used to estimate the motion of the echocardiogram sequence.Through dynamic tracking of the region of interest of the myocardial wall,the accuracy of the motion curve of each myocardial wall segment is guaranteed,and then the myocardial wall motion is vector decomposed,focus on taking the radial vector that is consistent with the motion direction of the myocardial wall for analysis,and finally get the motion information of each segment of the myocardial wall,which is expected to be used for the quantitative evaluation of the myocardial wall segment motion abnormalities,and assist the clinical rapid diagnosis.