Skip to main content

EMG-based characterisation of oscillatory neuromuscular drives

Final Activity Report Summary - EMG-CNMD (EMG-based characterisation of oscillatory neuromuscular drives)

To fully understand how muscle functions in healthy and diseased conditions, it is important to know how the human nervous system controls muscle during voluntary contractions. It is becoming increasingly evident that motor units, the smallest controllable units of muscle comprising a single motoneuron and the muscle fibres that it innervates, are not controlled completely independently of one another, but exhibit a task-dependent coupling, indicating the presence of shared or common inputs to the motoneuron pool. Furthermore, this coherent or correlated activity is observed across a range of distinct frequencies. The origin and functional significance of such common oscillatory neural inputs is not yet known.

The aim of this project was, therefore, to improve our understanding of the origin and functional role of coupling between simultaneously active muscles during voluntary contractions. A series of experimental studies were conducted to examine coupling between electromyographic (EMG) signals recorded from the hand and forearm muscles during sustained fatiguing muscle contractions in 15 healthy human subjects. Coherent activity, indicating the presence of common presynaptic neural inputs, was observed between synergistic hand muscles at the frequency of physiological tremor (8-12 Hz), and within the beta (15-30 Hz) and gamma (35-60 Hz) frequency ranges. Furthermore, a significant increase in coherence in the beta and gamma ranges was observed following fatiguing muscle contraction. A series of experiments was then conducted to examine whether similar fatigue-related changes are observed in the elbow flexor muscles, which are known to have weaker cortical-muscular connections than the muscles of the forearm and hand. Preliminary results indicate that there is a similar increase in coherence between EMG signals from the biceps and brachioradialis muscles following fatiguing elbow flexion, although a different distribution of the coherent activity across the frequencies of interest was observed.

In addition to the experimental studies, computational modelling studies were conducted to help understand the effect of common motoneuron inputs on the surface EMG signal. Computational modelling provides a means of examining the effect of different parameters and variables on the EMG signal and on EMG coherence, in a manner that is not possible in vivo or in vitro. Coherence between simultaneously active motor units was examined in response to a range of different common motoneuron input signals. The results confirm that coherence provides a reliable measure of common inputs to the motoneuron pool, however they suggest that caution should be taken when interpreting differences in coherence observed between motor units with significantly different firing properties or when comparing data with coherence in different frequency ranges. A separate simulation study was conducted to examine the widely employed technique of rectification of the surface EMG signal prior to calculating the coherence between signals. These results indicate that the frequency at which coherent activity is observed is maintained following rectification, however, the magnitude of the coherence is altered. Finally, EMG signals were simulated to examine the role of the surface EMG electrode in averaging the potential at the skin surface. The simulation results and analysis illustrate that under most practical conditions, the surface electrode provides a measure of the average potential at the skin surface beneath the electrode.

In conclusion, this project has examined coupling across surface EMG signals during voluntary isometric muscle contractions. The results indicate an increase in coupling and synchronisation across synergistic muscles increases in response to muscle fatigue. In addition, the simulation studies provide insight into the physiological and physical mechanisms that shape the EMG signal and suggest means for optimising the information which may be extracted from the EMG signal in order to accurately estimate the parameters of common or shared neural inputs to muscle.