基于相关反馈算法的周围动脉血管壁超声应变成像

Strain Imaging of Peripheral Artery based on Ultrasound Video Images and Correlation-feedback Technique

提出利用血压自主波动引起的血管壁组织位移与应变和超声视频图像 ,研究非刚性、小位移和高稳健的相关反馈运动估计算法 ,并与基于梯度的光流算法和基于相关算法进行对比。仿真实验和人体实验的结果证实相关反馈光流算法在时效和精确性上均能满足应变成像的需要。对循环系统生理病理学 。

In this paper,using displacements and strains of vessel wall tissues caused by pulsatile blood pressure and ultrasound video image sequence,we studied the non rigid motion estimating algorithm by which minute displacement and high quality strain images could be attained.Other two motion estimation algorithms are also applied for ultrasound strain imaging of carotid,and furthermore the performances of these algorithms are evaluated.The results of simulation and human experiments demonstrate that the correlation feedback algorithm can meet requirements in accuracy and efficiency.The method has potential to become useful tool for the field of circulation physiology and pathology,especially for hemodynamics.