The Research Of 3D Imaging Myocardium Strain And Elastic Modulus

Early diagnosis and treatment of heart disease contribute to the improvement of human health condition. Modern medical imaging technology has enabled us to observe the cardiac motion situation non-invasively. Three-dimensional (3D) motion and deformation analysis of the heart is of important significance in diagnosing cardiac diseases. Strain is a measure of mechanical deformation, which can directly reflect the extent of regional myocardial contraction and relaxation without the impact of the movement of the heart as a whole. So it is a quantitative and effective evaluation of myocardial function. The myocardial elastic modulus, also known as Young's modulus, as one of the myocardial material parameters that is closely related with deformation ability of the heart since it is the mathematical description of an object or substance's tendency to be deformed. Because the material parameters are sensitive measures of the object intrinsic property changes, especially for myocardium tissue, the quantification of left ventricular elastic modulus also has significant implication in clinic diagnosis of cardiac diseases.Now commonly used imaging methods of myocardium strain and elasticity have some shortcomings. Ultrasound elastography only has 1D measurement; MRE is 2D imaging and requires an external vibration source; the imaging qualities are quite low. Therefore, this paper presents novel robust algorithms for 3D imaging myocardial strain and elastic modulus based on cardiac cine-MRI images and optimization theory.In order to extract information of myocardial deformation from a cardiac cine-MRI image sequences reliably, this paper proposes a novel method to reconstruct complete left ventricular motion information and to implement strain imaging of left ventricle using adaptive biomechanical model which is constructed with prior material parameters initially and can be updated simultaneously through the procedure, based on extended Kalman filter (EKF) parameter estimation and the framework of joint segmentation and motion tracking of the left ventricle.This paper also studied the robust algorithm of 3D elastic modulus imaging with MRI images. With motion information corrupted by considerable noise, least squares method, extended Kalman filter estimation, full-state derivative information (FSDI) method and noise-perturbed full state information (NPFSI ) methods, both based on H-infinity filtering estimation, are used to achieve the reconstruction of the distribution of elastic modulus of left ventricle. The mehods are evaluated.