发布时间:2018-11-29浏览次数:47
讲座题目 | Learning-based methods for infant brain images analysis (基于学习的婴儿脑图分析方法) | ||||
---|---|---|---|---|---|
主办单位 | 信息科学与工程学院 | ||||
讲 座 人 | 王利 | 讲 座人 情 况 | 境外 | 主持人 | 蔡志明院长 |
讲座类型 | 自然科学 | 讲座对象 | 全校师生 | 时 间 | 2018-11-29 14:00 |
地 点 | 信息学院C1-206 | ||||
讲 座 人 简 介 | 王利,男,博士,助理教授,任职于北卡罗来纳大学教堂山分校,IEEE Senior member;2010年6月毕业于南京理工大学计算机科学与技术学院,获博士学位;2010年7月至2015年7月,在北卡罗来纳大学教堂山分校做博士后;2015年7月至2016年7月在北卡罗来纳大学教堂山分校任职讲师; 2016年7月至今任职助理教授。多年来一直致力于研究婴幼儿大脑研究,包括分割、重建、早期诊断,获得NIH Career Award (K01)和NIH R01项目;在国内外学术刊物和国际会议上发表学术论文138余篇,其中被SCI检索60余篇。Google Scholar总引用3447次,h-index=30。担任IEEE Transactions on Image Processing, IEEE Transactions on Medical Imaging, IEEE Transactions on Biomedical Engineering, IEEE Transactions on Cybernetics等30多个期刊的论文审稿人。 | ||||
讲 座 主要内容 | Recent progress in infant MRI technology allows us to track the dynamic brain developmental trajectories in vivo during the first year of life, which can greatly increase our very limited knowledge on normal early brain development, and also provide important insights into early neurodevelopmental disorders, such as autism spectrum disorder and schizophrenia. However, the existing neuroimaging computational tools, which were mainly developed for older children and adult brains, are ill-suited for infant brain studies, due to great challenges in tissue segmentation and labeling, caused by the extremely low contrast, insufficient resolution, severe partial volume effects, and dynamic growth. In this talk, Wang will introduce deep learning-based methods for infant brain images analysis, including tissue segmentation of cerebrum and cerebellum, hippocampal subfield, and imaging-biomarkers early diagnosis of autism. |