Study On Shear-wave Elastography And Motion Estimation Based On Time-domain Cross-correlation Algorithm

The elastic changes of biological tissues are usually directly related to the pathological phenomena,ultrasound elastography is to obtain pathological phenomena from the elastic changes of tissues,and the displacement motion estimation of biological tissues plays an important role.In the application process of clinical,the obtained displacement accuracy which is not high enough or the effect of real-time imaging is not achieved will affect the clinical prediction that the position of the lesion area cannot be accurately obtained from the image.Therefore,how to improve the displacement accuracy to achieve the real-time imaging has become a hot spot in this field.In this paper,the displacement motion estimation of elastography based on ultrasonic shear wave is studied.Firstly,the ultrasonic signal is preprocessed by filter.The improved adaptive weighted median filtering is proposed to make the data smoother.Secondly,based on the processing of signals,the shear wave elastography based on time domain crosscorrelation algorithm is proposed to achieve higher precision displacement calculation results,further noise reduction is carried out at the same time.Finally,based on imaging before,an adaptive block matching search method based on square template is proposed to get better displacement motion estimation.To sum up,this article focuses on the following three areas:(1)An improved adaptive weighted median filtering algorithm is proposed to filter the unprocessed ultrasonic signals.The improved algorithm based on adaptive median filtering in this paper improves the SNR value by about 11% and reduces the RMSE value by about 9%.(2)Shear wave elastography based on cross-correlation algorithm is proposed in this paper.The cross-correlation relationship between two frames is used to find the correlation of the signal and estimate its displacement.In the processing of data,parabolic interpolation method and window weighting processing are used: on the one hand,parabolic interpolation can improve the accuracy value of tissue displacement calculation;on the other hand,window weighting can further effectively filter out noise effects.The experimental results show that the shear wave imaging algorithm can improve the image signal-to-noise ratio by about 36.9% and reduce the error value by about 31.4%.(3)A block-matching motion estimation algorithm based on square adaptive(SAS)is proposed to calculate the displacement imaging results,and compared with several block matching methods.The final experiment results show that the SAS algorithm obtains better SNR value and MSE value,and has better motion matching estimation effect than other block matching methods.