A Novel Framework to Seek Functional Connectivity
In this paper,ICA (independent component analysis) is introduced in order to construct a feature space from the resting-state fMRrs data.The projection from the data space into the feature space reveals the regional clustering of function,and the subset of the space reveals the correlation across functionally related brain regions.We introduce SOM (self-organizing map) to assess regional functional connectivity as well as GA (genetic algorithm) to extract a feature subset,in which regional connectivity across functionally related brain regions can be assessed.The results of our framework demonstrate that the brain function is regionally clustered.
functional magnetic resonance imaging resting-state fMRI independent component analysis genetic algorithm self-organizing map
Songshi Dai Shanan Zhu
College of Electrical Engineering,Zhejiang University,HZ,China
国内会议
长江2011国际医学影像物理和工程应用大会暨第六届中国医学影像物理学术年会
杭州
英文
206-212
2011-10-22(万方平台首次上网日期,不代表论文的发表时间)