Electrocorticographic dissociation of alpha- and beta-band activity in human sensorimotor cortex
Arjen Stolk1, Loek Brinkman2, Mariska van Steensel2, Erik Aarnoutse2, Robert T. Knight1, Frans Leijten2, Floris de Lange3, Ivan Toni3; 1University of California, Berkeley, 2Utrecht University, 3Donders Institute
Alpha and beta rhythmic activities over the sensorimotor cortex are prominent and functionally relevant [Brinkman et al., 2014; 2016]. However, it is unclear whether alpha and beta rhythms build on spatially overlapping neuronal ensembles, and whether those ensembles actually contribute to computing a forthcoming movement. Complicating the issue is the fact that rhythmic activity rides on top of concurrent power-spectral 1/f modulations, making it difficult to make robust claims involving truly oscillatory activity. Here we recorded neural data directly from the cortical surface (ECoG) of eleven epilepsy patients while they prepared goal-directed movements with either the left or right hand. We used irregular-resampling auto-spectral analysis to distinguish subject-specific rhythmic from 1/f components of the ECoG signal [Wen & Liu, 2016]. Alpha and beta rhythms showed effector-specific trial-by-trial modulation, and were both spatiotemporally correlated with high-frequency activity (60-140 Hz). However, alpha and beta rhythms differed in their cortical and functional properties. Sensorimotor alpha is maximal on the postcentral gyrus, with the majority of electrodes yielding predominantly somatosensory sensations of the upper limb following electrical stimulation. In contrast, sensorimotor beta is strongest on both pre- and postcentral gyri, at electrodes yielding both movements and somatosensory sensations following stimulation. Further, each rhythm exhibited unique, non-overlapping spatiotemporal patterns, with beta rhythmic activity closely tracking fluctuations in excitation:inhibition balance across sensorimotor cortex [Gao et al., 2017]. Together, these observations suggest that alpha and beta involve different neuronal ensembles and dissociable components of movement computation.
Topic Area: PERCEPTION & ACTION: Motor control