财大数统学院
下载中心
讲座 您现在所在的位置: 首页资源分类讲座

张小群教授讲座公告
来源:中南财经政法大学统计与数学学院    编辑:佚名    时间:2017-5-8    点击数:

    报  人:张小群(上海交通大学)

    报告时间:5月12号下午2:30—3:10

    报告地点:文波楼4楼会议室

    题    目:Redundancy and sparsity exploiting for Dynamic Image reconstruction and segmentation

    摘   Dynamic medical imaging, such as dynamic PET/MRI/Ultrasound, are useful for the characterization of physiology process and assistance for operation navigation. In this talk, I will illustrate how we make use of data redundancy and sparsity in dynamic images for two medical imaging application: dynamic parallel MRI (pMRI) reconstruction and ultrasound video segmentation.  For dynamic pMRI, the problem of the central importance is to acquire less data in k-space domain while maintaining the quality of reconstructed images. By exploiting the between-frame redundancy of dynamic parallel MRI data, we propose a temporal-spatial low rank matrix based model for coil sensibility auto-calibration and then whole data set reconstruction. Numerical experiments on several highly subsampled data demonstrate that the proposed approach outperforms other state-of-the-art methods for cablirationless dynamic pMRI reconstruction.  The another application is on  ultrasound videos region-of-interest segmentation, which is a challenging task due to low image quality and real-time computation requirement for surgery navigation. We tackle the problem by using wavelet frames and incorporating the noise statistics under a variational framework. The continuity and regularity of the moving boundary is effectively incorporated via weighted sparse regularization, without introducing a heavy computational burden. The overall method can be efficiently solved with recently-developed fast algorithms, making it useful in real-time clinical applications.


    主讲人简介:

    张小群, 上海交通大学数学系与自然科学研究院教授,特别研究员。在武汉大学获得本科硕士学位,法国南布列塔尼大学获得应用数学博士。2007年至2010年之间在美国加州大学洛杉矶分校任访问助理教授, 2011获得上海市浦江人才计划,2013获得教育部新世纪优秀人才称号,2015年入选中组部青年拔尖人才计划称号。 主要从事数据科学, 图像科学,医学图像处理,计算机视觉中的数学模型与计算方法的研究, 在SIAM系列,Inverse Problems,Math. Comp等国际顶级杂志发表论文30多篇,论文引用1700多次。