Social and non-social learning of fmri-based neurofeedback of the ACC: Effects on cognitive interference

We developed a new paradigm, in which social feedback is provided. We studied the effectiveness of social feedback on anterior cingulate cortex (ACC) activity and tested the psychological effects on cognitive interference. Healthy volunteers (n=18), trained on 3 days, were divided into 2 different learning groups; in one group social feedback was provided to learn to regulate ACC activity, the other group learned through standard non-social feedback. During social feedback, an avatar provided a positive facial expression (smile) when the level of activity in the ACC increased and it gradually became neutral when the activity decreased. During standard non-social feedback a moving green bar indicated the ACC activity. We conducted this study to test if social reinforcement can train control of localized ACC activity, if a transfer to sessions without feedback emerged and if a generalization of the training to a cognitive interference task takes place. Our preliminary results indicate that regulation of localized ACC activation was learned and transfer of learned ACC control to sessions without feedback emerged. This effect was stronger with social neurofeedback. Thus, social feedback improved learning to control localized ACC activity compared to standard feedback. Networks underlying the Simon effect were modulated at the ACC after neurofeedback training. Localized neurofeedback can selectively and specifically modulate networks underlying cognitive interference processes.