Resting-state Functional Connectivity Modulates Motor Learning and Consolidation

Keara Greene, BA

Massachusetts General Hospital – Clinical Research Coordinator
Greene_Keara poster

Scientific Abstract

Learned information is consolidated during sleep; however, it is not well understood how the brain stores this information from learning to sleep. Earlier results suggest that resting state functional connectivity (rsFC) changes may predict sleep-dependent consolidation. We use magnetoencephalography (MEG) and electroencephalography (EEG) to further probe this effect. We hypothesize that after learning a motor task, the storage of information may be reflected in rsFC changes specifically in motor-related mu and beta frequency bands. 

We acquired simultaneous MEG and EEG data from 24 healthy controls during a motor learning task, pre- and post-learning rest, a 90-minute post-learning nap and a post-nap test on the same task. All session time courses were source-reconstructed to a 448-parcel connectome using combined EEG and MEG sensor data, and rsFC was computed among all parcels as the correlations between pairwise orthogonalized envelopes. We then tested for 1) the difference in rsFC pre- and post-learning, 2) correlations between the difference in rsFC pre- and post-learning and both learning during training and overnap improvement, and 3) correlations between pre-learning rsFC and both learning during training and overnap improvement.

While there were no significant changes from pre- to post-learning rsFC, increases from pre- to post-learning rsFC in cross-frequency and/or within-frequency was significantly positively correlated with both overnap improvement (r>.61,p<.001) and learning (r>.67, p<.001). Additionally, pre-learning rsFC was significantly negatively correlated with learning (r<-.7, p<.001).

The lack of significant learning-induced rsFC changes together with the correlations with post-learning rsFC increases indicates that small rsFC modulations have substantial effects on both learning and memory consolidation. Conversely, the negative correlations between rsFC and learning indicate that too strong coupling at rest may be detrimental to learning. While this study was carried out on healthy participants, we plan to expand it to patient populations.

Live Zoom Session – March 9th

research Areas


Keara D. Greene, BA, Martin Sjøgård, PhD, Dimitrios Mylonas, PhD, Sheraz Khan, PhD, Matti Hämäläinen, PhD, Robert Stickgold, PhD, Dara S. Manoach, PhD

Principal Investigator

Dara S. Manoach, PhD