Research On State Estimate Applied To Onmi-directional M-mode Cardiography

The heart is the most important organs of the human body.Meanwhile,a considerable amount of dynamic information about heart movement is hidden in high-level echocardiography images.Therefore,it's important to detect dynamic information of heart beating in the medical diagnosis.The LEJ-2 omni-directional M-mode echocardiography system offers a access to detect the dynamic information of the heart.The clinical studies have shown that it has auxiliary function to the early diagnosis of the cardiac disease.Supported by the National Natural Science Foundation of china(61471124)and the EU's Seventh Framework Programme(FP7),we focus on research of LEJ-2 omni-directional M-mode echocardiography system's motion curve detection,which had some disadvantages such as unsupervised and lack of stability.At the same time,the new issues of the accurate extraction of dynamic information raised in the clinical application,as well as a new algorithm is proposed to improve the original system,Main contents of this article are presented as follows:First,it introduced the algorithm of each step which LEJ-2 omni-directional M-mode echocardiography system used.Then,it analyzed the disadvantage of traditional edge detection which was applied to the system with the improved algorithms,such as incomplete information extraction and noise superposition problem.At the same time,the method which based on state estimation is proposed for the omni-directional M-mode echocardiography system.It introduced the advantage and the process of method.Second,an algorithm for the omni-directional M-mode echocardiography edge region extraction has been designed on the paper.The algorith is used in the coarse positioning.There are some features in omni-directional M-mode echocardiography,such as strong interference edge which is difficult to distinguish.At the same time,it need not catch the accuracy result on this step.Therefore,it proposed that built the model of ant colony Algorithm on omni-directional M-mode echocardiography and got the result by ant colony algorithm.In order to improve the initial search capabilities and identify number of the movement area,the algorithm combined annealing algorithm and used multiple type of pheromones.It prepared for status information extraction base on the state estimation.Third,an algorithm for the omni-directional M-mode echocardiography status information extraction base on the Kalman Filter has been designed by the author.it analyzed the characteristics and frame of the various modifications of the Kalman Filter,and selected the most appropriate algorithm which is fading Kalman Filter to extract the accuracy information.Then,it created the model based on the features of cardiac motion,and extract the position,speed and acceleration of cardiac.Experimental results show that it can obtain more smoother speed curve and acceleration curve,and the result is more matching the law of human heart motion and more accurate.Based on the related technologies of the detection of omni-directional M-mode echocardiography's motion curve and optimization for the detection of the heart's dynamic information,a new method which is based on the state estimation was proposed in this thesis.Multiple experimental resulted show that the final acceleration curve obtained by this method is more accurate than traditional method.It was presented for the future research.