The fMRI data are analyzed with BrainVoyager QX 1.9 (Brain Innovation, Maastricht, The Netherlands). Preprocessing steps include inter-slice time correction, motion correction, linear detrending, spatial alignment with the anatomical scans, warping into Talairach space, and spatial smoothing at 6 mm FWHM. Then, spatial independent component analysis (ICA) is applied to the fMRI single-subject data to delineate independent spatio-temporal patterns of brain activity [McKeown at al., 1998; Mantini et al., 2007b, 2009]. Data reduction by means of principal component analysis (PCA) with 60 dimensions is performed to retain at least 99.9% of the total variance. Independent components (ICs) are estimated by means of the FastICA algorithm [Hyvärinen et al., 1999]. Each fMRI-IC consists of a waveform and a spatial map: the waveform corresponds to the time-course of the specific pattern; the intensity of this activity across voxels is expressed by the associated spatial map. In order to perform group-level analysis, the ICs estimated from each subject are clustered by means of the self-organizing group ICA (sogICA) method [Esposito et al., 2005], selecting the number of clusters equal to that of the
