期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2015
卷号：4
期号：3
页码：1762
DOI：10.15680/IJIRSET.2015.0403
出版社：S&S Publications
摘要：Brain interaction patterns are very complex and FMRI provides a prospective to stud y the functioning of brain and identifying the brain state.The FMRI data are time series of 3 -dimensional volume images of the brain. The data is traditio nally analyzed within a mass-univariate framework essentially rel ying o n classical inferential statistics. Handling of feature selection and clustering is a complicated process in interaction patterns of brain datasets. To understand the complex interaction patterns amo ng brain regions system proposes a novel clustering techniq ue. System models each subject as multivariate time series, where the single dimensions represent the FMRI signal at different anatomical regions. In the Proposed paper, three algorithms such as FSS, IKM and Dimension Ranking Algori thm are analyzed.
