| Cardiovascular diseases are the main cause of human deaths.According to the World Health Organization(WHO),approximately 6.5 million people have died of cardiovascular disease worldwide in the past 30 years,which takes approximately one-third of all death cases.It is of great significance for the prevention of heart diseases to quantitatively evaluate heart function at early stages.Quantitative evaluation of cardiac function usually relies on various medical imaging technologies.Among these technologies,echocardiography has been used most widely.It is cheap,portable,and free of ionizing radiation.It also has high time resolution and is suitable for dynamic study on cardiac motion.2D/3D Speckle Tracking Echocardiogram(2D/3D STE)is an important technique to quantify myocardial function.It obtains the displacement estimation and strain parameters of heart by tracking the speckles in the echocardiogram frame by frame.In this way it helps the diagnosis and prevention of heart disease Compared with tissue Doppler imaging,the speckle tracking has no angle-dependent limitations and has the advantage of quantifying local tissue deformation.It is widely used to characterize the multi-dimensional motion of the heart and has important clinical value in evaluating myocardial dysfunction(such as infarction and ischemia).Strain imaging with 2D STE has been successfully utilized to quantify myocardial function.However,since the motion of the heart and the speckle are essentially 3D processes,it requires 3D imaging to accurately track the speckle.3D STE tracks myocardial deformation through actual 3D observations rather than geometric assumptions,which can overcome some inherent limitations of 2D STE.For example,it can avoid errors caused by heart rhythm variability when 2D STE performs multiple acquisitions,and avoid speckle tracking failure when the speckle moves out of the imaging scope.Additionally,all the 3D strain parameters of the left ventricle can be calculated through a single volume of dataset.In conclusion,3D STE is theoretically superior to 2D STE and traditional echocardiographies.2D/3D echocardiographic speckle tracking methods could be divided into three categories:methods based on optical flow,methods based on elastic image registration,and block-matching methods.Speckle tracking based on optical flow and elastic image registration needs to be based on high-quality images,and is accompanied by complex calculations,which is poor in clinical practicability.The block-matching speckle tracking methods are the mainstream in the field of 3D STE imaging technology because of their simple implementation and great clinical practicability.However,the traditional block-matching methods are difficult to achieve high-precision when tracking myocardial tissue motion.In addition,the adaptive characteristics and robustness of previous methods are far from satisfactory.To tackle the low computational efficiency or low tracking accuracy of previous methods,this thesis proposes a fast 3D echocardiographic speckle tracking method with sub-pixel accuracy.The method is based on block matching 3D sub-pixel speckle tracking,and is denoted as SBM method.The SBM method has great adaptive characteristics and robustness,and can realize fast speckle tracking with 1/16 sub-pixel accuracy.This thesis also performs simulations on the STRAUS(3D Strain Assessment in Ultrasound)dataset with the SBM method,and compares SBM to another four methods,which are based on optical flow,elastic registration or block matching.The results illustrate the effectiveness and great tracking accuracy of the SBM method,and also reveal that the SBM method has high clinical value and potential for application. |
PDF Full Text Request