Although neuroimaging techniques have provided vital insights into the anatomical regions involved in motor learning, the underlying changes in temporal dynamics are not well understood. Using magnetoencephalography and Hidden Markov Modelling to model the dynamics of neural oscillations on data-adaptive time-scales, we detected specific changes in movement-related sensorimotor β-activity during practice of a self-paced sequential visuo-motor task. The behaviourally-relevant neural signature generalised to another motor task, emphasising the centrality of β-activity in motor plasticity.
MainPlasticity refers to the ability of organisms or cells to alter their phenotype in response to changes. One fundamental example of human plasticity which may serve as an exemplar is learning a new motor skill. (f)MRI and PET studies have provided critical insights as to which brain regions are involved in motor learning 1,2 .However, the temporal dynamics underpinning motor skill acquisition are not well understood. To date, only electrophysiological recordings provide the temporal precision required to capture the rich temporal dynamics of neural activity.Movement-related changes in cortical sensorimotor β-activity (13-30 Hz, perimovement 1 power decrease and post-movement power increase) have been shown to play a key role in movement 3 , making them, putatively, a neural marker for motor learning. However, their role in motor plasticity remains largely unknown.So far human studies have provided only indirect, and partly contradictory, evidence for behaviourally-relevant learning-induced changes in sensorimotor β-activity 4-7 .While animal studies have identified somatostatin-expressing interneurons in the motor cortex as a key regulator of motor learning 8 , which, in the primary visual cortex, have been related to preferentially drive β-oscillations 9 , direct evidence for changes in β-activity in sensorimotor areas driving motor learning has yet to be established.1 We use the term ‚peri-movement' to refer to the time from movement onset to movement offset.All rights reserved. No reuse allowed without permission.(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint . http://dx.doi.org/10.1101/345421 doi: bioRxiv preprint first posted online Jun. 12, 2018; Zich et al., Motor learning shapes temporal activity in human sensorimotor cortex -Manuscript 3 Here, we used MEG to directly capture practice-induced functional changes in temporal dynamics during learning of a self-paced, and thus more naturalistic, sequential visuo-motor task ( Supplementary Fig.1 Supplementary Fig.2b) correlated significantly with behavioural improvement.Conventional Fourier-based, wavelet analysis relies on pre-specification of a wavelet length parameter, which constrains the time-scale over which temporal dynamics can evolve. Whilst this provides a good representation of tasks with consistent timing over trials it is likely to be too ...
