Cortical activation related to arm-movement combinations

Recent studies support the long-standing hypothesis that continuous arm mov ements consist of overlapping, discrete submovements. However, the cortical activation associated with these submovements is unclear. We tested the hy pothesis that electroencephalography (EEG) activity would more strongly cor respond to the particular combinations of muscle electrical activity, the i ndependent components (ICs) of surface electromyography (EMG), than the sur face EMG from individual muscles alone. We examined data recorded from two normal subjects performing sustained submaximal contractions or continual, unpaced repetitive movements of the arm. Independent component analysis (IC A) was used to determine the ICs of the multichannel EMG recordings (EMGICs ). ICA was also used to calculate the coupling between the simultaneously r ecorded EEG and the EMG from a single muscle (Subject 1) or the EMGICs (Sub ject 2). The EMGICs were either tonic or phasic, The significant couplings between the EEG and the EMGICs were different for each EMGIC, The distribut ion on the scalp of the coupling between the EEG and tonic EMGICs and those of the single-muscle EMG were similar and followed topographic patterns in sensorimotor regions. Couplings between the EEG and phasic EMGICs were bif rontal, lateral, and bioccipital and were significantly stronger than the c oupling between a single muscle's EMG and the EEG (p < 2 x 10(-5)) Or anoth er EMG combination derived from principal component analysis. These prelimi nary results support the notion that electrophysiological cortical activati ons are more significantly related to the ICs of muscle activations than to the activations of individual muscles alone. (C) 2000 John Wiley & Sons, I nc.