Learning of new associations invokes a major change in modulations of cortical beta oscillations in human adults

Large-scale cortical beta (beta) oscillations were implicated in the learning processes, but their exact role is debated. We used MEG to explore the dynamics of movement-related beta-oscillations while 22 adults learned, through trial and error, novel associations between four auditory pseudowords and movements of four limbs. As learning proceeded, spatial-temporal characteristics of beta-oscillations accompanying cue-triggered movements underwent a major transition. Early in learning, widespread suppression of beta-power occurred long before movement initiation and sustained throughout the whole behavioral trial. When learning advanced and performance reached asymptote, beta-suppression after the initiation of correct motor response was replaced by a rise in beta-power mainly in the prefrontal and medial temporal regions of the left hemisphere. This post-decision beta-power predicted trial-by-trial response times (RT) at both stages of learning (before and after the rules become familiar), but with different signs of interaction. When a subject just started to acquire associative rules and gradually improved task performance, a decrease in RT correlated with the increase in the post-decision beta-band power. When the participants implemented the already acquired rules, faster (more confident) responses were associated with the weaker post-decision beta-band synchronization. Our findings suggest that maximal beta activity is pertinent to a distinct stage of learning and may serve to strengthen the newly learned association in a distributed memory network.

Automated summary

This study used MEG technology to investigate the dynamics of movement-related beta oscillations in the learning process. The results showed that beta oscillations undergo significant changes as learning progresses. In the early stages of learning, beta power is suppressed before movement initiation, while at later stages, there is an increase in beta power after correct motor response. These post-decision beta oscillations can predict trial-by-trial response times. This suggests that beta activity is crucial for a specific stage of learning and may be involved in strengthening newly learned associations in the brain.