Since the EEG recordings collected in the MRI scanner were contaminated by imaging, BCG, and ocular artifacts, we applied artifact attenuation techniques to obtain signals that could be useful for EEG–fMRI integration. Given that EEG and fMRI were recorded simultaneously, we used the slice-based AAS method for artifact imaging subtraction (Gonçalves et al., 2007), thus avoiding a possible P300 attenuation induced by the classical gradient-based AAS method (Allen et al., 2000). The residual imaging contamination in our EEG data, as well as the BCG and ocular artifacts, were then removed by means of an ICA-based procedure, which was already tested and used in our previous EEG–fMRI studies (Mantini et al., 2007a,b).
