Spatiotemporal group ICA applied to fMRI datasets

Research Area: Research Year: 2008
Type of Publication: Article
Authors: Kohler, C H; Keck, I; Gruber, P; Lie, C H; Specht, Karsten; Tome, A M; Lang, E W
Abstract:
Exploratory data analysis techniques such as independent component analysis (ICA) do not depend on a priori hypotheses and are able to detect unknown, yet structured spatiotemporal processes in neuroimaging data. We present fMRI data of two different subject-groups (young and old), which performed a modified Wisconsin Card Sorting Test (WCST). Spatiotemporal ICA and SPM-generated brain maps of the subject data are compared. For the group analysis a singular value decomposition approach was used. Spatiotemporal ICA reveals a frontoparietal network being activated while subjects performed different variants of the WCST. Contrary to the SPM analysis, ICA analysis revealed significant differences between young and old subjects as well as significant within-group differences.While young subjects showed with increasing task demands (A>>B>>C) increasing activation of the right lateral prefrontal cortex and of the medial orbito-frontal cortex, old subjects showed no such gradient in activation pattern and appeared to be more distributed.
Digital version