Resampling as a cluster validation technique in fMRI

Exploratory, data-driven analysis approaches such as cluster analysis, prin cipal component analysis, independent component analysis, or neural network -based techniques are complementary to hypothesis-led methods. They may be considered as hypothesis generating methods. The representative time course s they produce may be viewed as alternative hypotheses to the null hypothes is, ie, "no activation." We present here a resampling technique to validate the results of exploratory fuzzy clustering analysis. In this case an alte rnative hypothesis is represented by a cluster centroid. For both simulated and in vivo functional magnetic resonance imaging data, we show that by pe rmutation-based resampling, statistical significance may be computed for ea ch voxel belonging to a cluster of interest without parametric distribution al assumptions. (C) 2000 Wiley-Liss, Inc.