A study on cortico-muscular coupling in finger motions for exoskeleton assisted neuro-rehabilitation

In this paper our objective is to analyze the cortico-muscular coupling for hand finger motion and its possible use in the control of an exoskeleton based neurorehabilitation system for stroke sufferers. Cortical activity alone is often not sufficient to reliably control a device such as an exoskeleton and hence, our focus is to ascertain and analyze the connectivity between the motor cortex and forearm muscles, controlling the fingers, in terms of coherence between electroencephalogram (EEG) and electromyogram (EMG) signals. We have analyzed the signals separately for three different kinds of exercises consisting of passive motion of fingers using exoskeleton, active motion without any assistance, and motor imagery of the same movements. Four out of six healthy subjects who participated in the experiments have shown significant (p<;0.01) coherence for active finger motion which is well distinguished from the rest state. The EEG analysis resulted in average accuracy of 69.17% for passive finger motion with exoskeleton, 71.25% for active finger motion, and 67.92% for motor imagery, in detecting the volitional intention of the subjects to move their fingers. These results support that EEG-EMG coherence along with EEG analysis has the potential to make a more effective neurorehabilitation system for finger movement restoration of stroke sufferers.