Phasic and tonic coupling between EEG and EMG demonstrated with independent component analysis

The authors describe a method for demonstrating the tonic and phasic coupli ngs between suitably time-aligned surface eletromyographs (sEMGs) and the s imultaneously recorded EEGs. The method, based on independent component ana lysis, was applied to data recorded from two normal subjects performing sus tained submaximal contractions or continual repetitive movements of the arm . Augmented datasets, consisting of the EEG and either the sEMG from a sing le muscle (subject 1) or a combination of sEMGs from several muscles (subje ct 2), were analyzed with independent component analysis to determine the E EG/sEMG coupling. Each derived coupling consisted of a spatial distribution on the scalp and a waveform representing an EEG channel combination coacti vating with the sEMG. The combinations of sEMGs, derived by applying indepe ndent component analysis to the simultaneous sEMG recordings from several m uscles to create sEMG independent components (ICs), were either tonic or ph asic with differing periods of activation. The topographic distributions on the scalp of the couplings between the EEG and sEMG ICs were different for each sEMG IC. The spatial distributions of the couplings between tonic sEM G ICs or single-muscle sEMGs and the EEG followed topographic patterns in s ensorimotor regions. Phasic couplings were bifrontal, lateral, and bioccipi tal. Calculation of coherence between the sEMG ICs and calculated EEG combi nations agreed well with the frequency spectra of the independent component analysis-derived coupling waveforms. These preliminary results demonstrate that detection of both the tonic and phasic coupling between the sEMG and the EEG is possible when monitoring unpaced proximal arm movement. This may thus be a practical means of exploring the dynamic cortical/muscle relatio nships in subjects unable to perform fine finger movements, such as patient s recovering from stroke.