Research On Motion Tracking And Displacement Estimation Algorithm For Ultrasonic Elastic Imaging Based On Particle Swarm Optimization

In ultrasound elasticity imaging, treatment of biological tissue for motion tracking measurement and displacement estimation is a very important process, that is they has always been a hot research topic in the field. According to displacement estimation problem, researchers have proposed and phase zero cross-correlation algorithm search classic algorithms, etc. But these imaging algorithm of computing performance can meet the real-time and clinical application of ultrasound elasticity imaging. Based on particle swarm optimization is proposed in this thesis, therefore, ultrasound elasticity imaging tracking and displacement estimation algorithm to improve the quality of ultrasound elasticity imaging, and it can satisfy the real-time imaging.This thesis first describes the estimated by particle swarm optimization (each to estimate point displacement process in detail, and then based on GPU computing platform, GPU parallel implementation of the proposed method, finally through the simulation experiment data with traditional cross-correlation algorithm is analyzed. The simulation results show that the method based on particle swarm optimization is better than traditional cross-correlation algorithm can accurately estimate organization movement situation, make the elastic figure is of high quality, at the same time GPU parallel implementation of the method effectively improves the calculation speed, can satisfy the requirement of real-time ultrasound elasticity imaging. To sum up, this thesis mainly do the following work:1. Put forward the ultrasound elasticity imaging algorithm based on particle swarm optimization, to estimate the longitudinal displacement of ultrasound elastography. With the simulation experiment proves that the method based on particle swarm optimization is better than traditional cross-correlation method can accurately estimate organization movement situation, can get high quality elastic graph.2. Against proposed based on particle swarm optimization algorithm of ultrasound elastography motion tracking and displacement estimation method is further optimized, in order to meet the real-time imaging, studied the GPU technology based on CUDA platform, using the GPU multi-core features, first of all, based on CUDA parallel PSO algorithm to carry on the design and implementation, and then to estimate the displacement of ultrasound elastography of parallelism is analyzed, parallel framework is given, at last, through repeated based on particle swarm optimization is verified by the experiments of ultrasound elastography movement track and parallelization of displacement estimation method and the particle swarm optimization algorithm based on GPU and CPU based method to calculate the comparative analysis of time, get a different speed ratio, to verify the efficiency of this method.