Real-time fMRI neurofeedback is an increasingly popular neuroimaging technique that allows an individual to gain control over his/her own brain signals, which can lead to improvements in behavior in healthy participants as well as to improvements of clinical symptoms in patient populations. However, a considerably large ratio of participants undergoing neurofeedback training do not learn to control their own brain signals and, consequently, do not benefit from neurofeedback interventions, which limits clinical efficacy of neurofeedback interventions. As neurofeedback success varies between studies and participants, it is important to identify factors that might influence neurofeedback success. Here, for the first time, we employed a big data machine learning approach to investigate the influence of 20 different design-specific (e.g. activity vs. connectivity feedback), region of interest-specific (e.g. cortical vs. subcortical) and subject-specific factors (e.g. age) on neurofeedback performance and improvement in 608 participants from 28 independent experiments. With a classification accuracy of 60% (considerably different from chance level), we identified two factors that significantly influenced neurofeedback performance: Both the inclusion of a pre-training no-feedback run before neurofeedback training and neurofeedback training of patients as compared to healthy participants were associated with better neurofeedback performance. The positive effect of pre-training no-feedback runs on neurofeedback performance might be due to the familiarization of participants with the neurofeedback setup and the mental imagery task before neurofeedback training runs. Better performance of patients as compared to healthy participants might be driven by higher motivation of patients, higher ranges for the regulation of dysfunctional brain signals, or a more extensive piloting of clinical experimental paradigms. Due to the large heterogeneity of our dataset, these findings likely generalize across neurofeedback studies, thus providing guidance for designing more efficient neurofeedback studies specifically for improving clinical neurofeedback-based interventions. To facilitate the development of data-driven recommendations for specific design details and subpopulations the field would benefit from stronger engagement in Open Science and data sharing.

Rewards can be parsed into a motivational component ('wanting'), which relies mainly on the brain's dopaminergic system, and a hedonic component ('liking'), which relies on the opioidergic system. The observation of animal facial and behavioral reactions to rewards (e.g. pleasant tastes) has played a crucial role for our understanding of the neurochemical bases of reward processing. In adult humans, however, implicit facial reactions to reward anticipation and consumption are rarely reported, and the role that the dopaminergic and opioidergic systems play in human facial reactions to rewards remains largely unknown. It is also unclear, if human facial reactions to different types of rewards have the same neurochemical basis. To answer these questions, we conducted a study using a randomized, double-blind, betweensubject design in which 131 volunteers (88 females) received orally either the D2/D3 receptor antagonist amisulpride (400 mg), the non-selective opioid receptor antagonist naltrexone (50 mg), or placebo. Explicit (ratings and physical effort) and implicit (facial EMG) reactions to matched primary social and nonsocial rewards were assessed on a trial-by-trail basis. Sweet milk with different concentrations of chocolate flavor served as nonsocial food reward. Gentle caresses to the forearm, delivered by the same-sex experimenter at different speeds, served as social reward. Results suggested 1) reduced wanting of rewards after administration of both dopamine and opioid receptor antagonists, compared to placebo, as indicated by less physical effort produced to obtain the announced reward and increased negative facial reactions during reward anticipation; 2) reduced liking of rewards only after administration of the opioid receptors antagonist, compared to placebo, as indicated by reduced positive and increased negative facial reactions during and following reward consumption. Most drug effects were either stronger or restricted to food trials, suggesting that wanting and liking of both social and nonsocial rewards may only partially share the same neurochemical brain substrates. The results are in line with the distinction of wanting and liking by current theories of reward, and underline the importance of assessing implicit facial reactions when conducting research on reward processing in adult human participants.