A New Hierarchical Clustering Analysis and the Application in Localization of Brain Activities  

作　　者：YUAN Hong CHEN Huafu YAO Dezhong CHEN Wufan 

机构地区：[1]School of Life Science and Technology, School of Applied Mathematics, University of Electronic Science and Technology of China, Chengdu 610054, China [2]School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China [3]Key Laboratory of Personality and Cognition, Southwest University, Ministry of Education of China, Chongqing 400715, China

出　　处：《Chinese Journal of Electronics》2006年第4期679-681,共3页电子学报（英文版）

摘　　要：Functional magnetic resonance imaging （fMRI） has emerged as a useful tool in the study of brain function, but the data process of fMRI is also unsolved. In this paper, proposed is a new hierarchical clustering analysis presented to localize brain function based on the idea of function group of human brain and a new spatial-temporal distance analysis. This method is that a Neighborhood correlation （NC） is utilized at first to get a primary imaging to choose the potential activation pixels and discard nonactivation pixels to decrease the complexity of hierarchical algorithm. Then the new hierarchical clustering analysis is implemented to foci brain activation. The simulation test and the localization of brain activities in the vivo fMRI data show that it provides more accuracy and faster speed in detecting weak fMRI signals.

关 键 词：Functional magnetic resonance imaging （BMC） Clustering analysis Spatial-temporal distance Near-local-correlation. 

分 类 号：O44[理学—电磁学]