FMRI images were analyzed using a widely available software package SPM5 (Wellcome Department of Imaging Neuroscience, University College, London, UK) running in Matlab (MathWorks, Natick, MA, USA). Initially, blood oxygen level dependent (BOLD) fMRI data were corrected for motion in SPM5 using an intra-run realignment algorithm that uses the first image as a reference. A criterion of 2 mm of head motion in any direction was used as an exclusionary criterion. The realigned images were then normalized to an EPI template in Montreal Neurological Institute (MNI) stereotactic space. Normalized images were re-sampled into 2 mm cubic voxels and then spatially smoothed using an isotropic Gaussian kernel with 8 mm full width at half maximum (FWHM). Global scaling was not used, high-pass temporal filtering with a cut-off of 128s was applied, and serial autocorrelations were modeled with an AR(1) model in SPM5. Statistical parametric images were calculated individually for each subject and each task, using a general linear model (Friston et al., 1995a; Friston et al., 1995b). These images were subsequently entered into second level model, subjected to a voxel-wise contrast
