Interaction K-means Clustering for Finding Similar Interaction Patterns among Brain Regions

Abstract: Brain is an important part, which controls all the functions of the human body. Brain activity is the only resource to understand whether any brain disorders exist or not .Functional Magnetic Resonance Imaging (FMRI) paves the way to study about the brain functions. But the information content from that is large in volume and complex and so the data requires some effective and efficient data mining techniques. To understand the complex interaction patterns among brain regions novel clustering technique is proposed.The objective is to assign objects exhibiting a similar intrinsic interaction pattern to common cluster. Based on this novel cluster notion, interaction K-means (IKM) is proposed, an efficient algorithm for partitioning clustering. IKM simultaneously clusters the data and discovers the relevant cluster-specific interaction patterns. The results on two real FMRI studies demonstrate the potential of IKM to contribute to a better understanding of normal brain function and the alternations characteristic for psychiatric disorders.

Keywords: Clustering, fMRI data, Interaction patterns, Multivariate time series.