A Maximal-Correlation Approach Using Ica for Testing Functional Network Connectivity Applied to Schizophrenia.

Conference PaperinProceedings / IEEE International Symposium on Biomedical Imaging: from nano to macro. IEEE International Symposium on Biomedical Imaging · January 2007with3 Reads
DOI: 10.1109/ISBI.2007.356890 · Source: DBLP
Conference: Proceedings of the 2007 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, Washington, DC, USA, April 12-16, 2007


    There has been a growing interest in analyzing brain activation differences between patients and controls by studying resting-state fMRI brain networks. Functional connectivity of the resting brain has been studied by analyzing correlation differences in time courses among seed voxels, regions, or volume of interest with other voxels of the brain in patients versus controls. Spatial differences have also been analyzed among component maps derived from independent component analysis (ICA) in patients with schizophrenia and in healthy controls. However, the relationship among ICA component time courses, (which we define as functional network connectivity), has not been studied. We propose a novel technique to determine FNC applied to schizophrenia which does not rely on the time series of specific brain voxels or regions of interest and instead focuses upon the connectivity between functional networks (components) estimated from ICA using maximal correlation between component time series.