Functional images were reconstructed using an iterative algorithm 41, which is more robust against image distortions caused by off-resonance effects than conventional methods. Subject head motion and slice-acquisition timing were corrected using the FSL 4.0 analysis tools library (Analysis Group, FMRIB, Oxford, UK) 42. Analysis of estimated motion parameters confirmed that overall head motion within each run did not exceed 2 mm translation or 2 degree rotation in any direction. All remaining image processing was completed using statistical parametric mapping SPM2 package (Wellcome Institute of Cognitive Neurology, London, UK). Functional images were spatially normalized to a standard stereotactic space as defined by the Montreal Neurological Institute (MNI). A 6 mm full width half maximum (FWHM) Gaussian spatial smoothing kernel was applied to improve signal to noise ratio and to account for individual differences in anatomy.
