Motor cortical beta transients delay movement initiation and track errors

ABSTRACT Motor cortical activity in the beta range (13-30 Hz) is a hallmark signature of healthy and pathological movement, but its behavioural relevance remains unclear. Recent work in primates and human sensory cortex suggests that sustained oscillatory beta activity observed on average, may arise from the summation of underlying short-lasting, high-amplitude bursts of activity. Classical human movement-related event-related beta desynchronisation (ERD) and synchronization (ERS) may thus provide insufficient, non-dynamic, summaries of underlying focal spatio-temporal burst activity, limiting insight into their functional role during healthy and pathological movement. Here we directly investigate this transient beta burst activity and its putative behavioural relevance for movement control, using high-precision magnetoencephalography (MEG). We quantified the subject-specific (n=8), trial-wise (n>12,000) dynamics of beta bursts, before and after movement. We show that beta activity on individual trials is dominated by high amplitude, short lasting bursts. While average beta changes generally manifest as bilaterally distributed activity (FWHM = 25mm), individual bursts are spatially more focal (FWHM = 6 mm), sporadic (1.3 −1.5/s), and transient (mean: 96 ms). Prior to movement (the period of the classical ERD), the timing of the last pre-movement burst predicts movement onset, suggesting a role in the specification of the goal of movement. After movement (the period of the classical ERS), the first beta burst is delayed by ~100ms after a response error occurs, intimating a role in error monitoring and evaluation. Movement-related beta activity is therefore dominated by a spatially dispersed summation of short lasting, sporadic and focal bursts. Movement-related beta bursts coordinate the retrieval and updating of movement goals in the pre- and post-movement periods, respectively.