Combining hypothesis-generating and hypothesis-testing tools during analyses of multisensory fMRI data
Naumer, M. J.1, van den Bosch, J. F.1, Wibral, M.2, Kohler, A.3, Singer, W.3, Kaiser, J.1, van de Ven, V.4, and Muckli, L.5
1Institut für Medizinische Psychologie, Goethe Universität, Frankfurt am Main; 2Brain Imaging Center, Goethe Universität, Frankfurt am Main; 3Max Planck Institut für Hirnforschung, Frankfurt am Main; 4Faculty of Psychology, Maastricht University, The Netherlands; 5Department of Psychology, University of Glasgow, United Kingdom
In this study, we aimed at mapping the network of brain regions that were functionally connected during object-related audio-visual (AV) integration. To this end, we used spatial independent component analysis (sICA), a multivariate, data-driven analysis technique that decomposes an fMRI dataset into spatially independent components. The resulting components were classified by the functional profile of their time course in a general linear model (GLM) of the stimulation time course, selecting three components of interest: visual, auditory, and audio-visual. Regions-of-interest (ROIs) were defined as clusters of voxels which weighted significantly to at least two of these, in other words, these were regions of overlap. ROI-restricted GLM analyses on the independent data of a second experiment showed robust AV integration effects according to the max-criterion (i.e., 0<A<AV>V>0) in this ICA-defined bilateral cortical network (consisting of pSTS, VOT, PPC, and PFC regions), thus demonstrating the sensitivity and value of the proposed analysis approach.