Prof. Weichuan Yu

Hong Kong University of Science and Technology (HKUST)

A Coupled Filtering Method for Solving the Feature-Motion Decorrelation Problem in Ultrasound Speckle Tracking

In ultrasound image-based tissue deformation analysis, feature-motion decorrelation causes feature tracking results fail to represent the underlying true tissue deformation. When tissue undergo large deformation, the relative estimation error can be as large as over 200%. We propose a new coupled filtering method to solve the feature-motion decorrelation problem. After explicitly modeling image variations caused by tissue deformation, we filter the image before tissue deformation and the warped image after tissue deformation with a pair of filters, respectively. We show through theoretical derivation that the two filtered images are identical to each other. Since the coupled filtering method keeps tissue deformation parameters, feature-based tracking is able to accurately estimate the parameters of underlying tissue deformation. Experiments in elastography studies show that our method is superior over existing methods, reducing the relative estimation error from over 200% down to 2-3%. This is a joint work with Tianzhu Liang, Ling Sing Yung, Chen Zhang and Philip Mok. Bio: Weichuan Yu is a Professor in the Department of Electronic and Computer Engineering at the Hong Kong University of Science and Technology (HKUST). He received his Ph.D. degree in Computer Vision and Image Analysis from University Kiel, Germany in 2001. He was a postdoctoral associate at Yale University from 2001 to 2004, and a research faculty member in the Center for Statistical Genomics and Proteomics at Yale University from 2004 to 2006 before joining HKUST in 2006. He is interested in computational analysis problems with biological and medical applications. He has published around 100 journal articles and referred conference papers on a variety of topics including bioinformatics, computational biology, biomedical imaging, signal processing, pattern recognition and computer vision. He was the recipient of the DAAD (German Academic Exchange Center) Fellowship, Biomedical Wellness Pioneer Award in recognition for the contribution to biomedical computing in SPIE Conference on Defense, Security and Sensing held in Baltimore in 2013, and the Best Paper Award in the Fourteenth Asia Pacific Bioinformatics Conference (APBC’16) held in San Francisco in 2016. He has been serving as an Associate Editor for BMC Bioinformatics since 2012 and an Associate Editor for Biology Direct since 2013.