时间：2012年12月19日下午1：30

地点：629会议室

报告人：张少霆博士

主持人：孙仕亮教授

报告简介：A national priority, through the National Institutes of Health and National Science Foundation, is to improve healthcare conditions. Following this context, biomedical image computing and anatomy/physiological systems modeling are effective and noninvasive approaches to aid medical doctors in accurate and reproducible diagnosis. For examples, 1) modeling cardiac motions and simulating blood-flow can help doctors analyze CT and/or tagged MRI data, and diagnose hypertrophy patients. 2) Modeling the joint kinematics using fluoroscopic X-ray images can benefit the investigation of injury prevention. 3) Modeling the strain of prostate using ultrasound-based elasticity-imaging can help to detect prostate cancers. These applications are particularly challenging, and at the core of these challenges lay the capability to effectively process these (3D) biomedical images and model the anatomy and physiological systems robustly and appropriately. This is critical to the accuracy of the diagnosis and the performance of the clinical analysis. In this talk, I will introduce several robust and scalable algorithms of image analysis and anatomy modeling, and I will also introduce the above-mentioned applications of computational diagnostics. This is collaboration with radiologists and clinical experts in the medical center of New York University, Columbia University, University of Pennsylvania and Brookhaven National Lab.

报告人简介：张少霆，美国Rutgers大学计算机系研究助理教授，于2005年在浙江大学软件学院获得学士学位，2008年3月在上海交通大学获得计算机软件与理论硕士学位，2012年1月在Rutgers大学计算机科学专业获得博士学位。张博士的主要研究领域包括医学图像分析，计算机视觉，机器学习等。在国内外知名期刊和国际会议上发表论文40余篇，发明专利6项，其中包括10余篇期刊(3篇Medical Image Analysis，一区，影响因子4.4)，以及16篇顶级会议文章(CVPR/ICCV/ECCV/MICCAI)。张博士在医学图像分析领域的相关工作于2010年获得MICCAI青年科学家奖(医学图像顶级会议)，并于2011年再次获得该奖提名。他近年来还作为核心成员申请并参与了若干美国国家科学基金以及美国国家卫生研究院的重点项目。