Contribution of cortico muscular synchrony to human movement and its potential use in neuroprosthetic control
The growing interest in the development of robust methods of intention detection for use in the activation and regulation of prosthetic devices and communication aids for people with motor impairments has led to the examination on synchronous rhythmic activity of the nervous system as a source of signals for use in the activation and control of neuroprosthetic devices. It has been reported that EMG activity during movement shows high spectral power and high intermuscular coherence in the 8-12Hz range. During posture, the EMG is modulated in the 15- 30Hz band and this activity is correlated to localised beta activation of the motor cortex measured by EEG. This thesis further investigates these task dependencies using multichannel EEG and EMG recording. Robust detection and interpretation of movement dependent features will create the prospects for efficient neuroprosthetic devices driven by patients activating brain or muscle components associated with the motor behaviour replicat ed by the prosthesis. Beta (15-30Hz) EEG power, intracortical coupling, corticomuscular coupling, and intermuscular coupling showed the most consistent task dependent features during posture. The cortical activity, underlying the beta corticomuscular coherence was localised in the area over the contralateral motor cortex. During movement the beta modulation was suppressed while agonist antagonist intermuscular coherence was replaced by 8-12Hz features likely to have a central origin. Changes in attention state and simultaneous cognitive activity did not affect the robustness of the corticomuscular and intermuscular features. This thesis provided useful insights on the involvement of central oscillatory activity in motor control and its manifestations in the periphery. Despite the not clear functional significance of the identified features, and the subject variability it was concluded that the examined signals and associated task dependent characteristics have a significant potential for use in neuroprosthetic control.