The surface electromyogram as an index of activity in a population of motor units
The nervous system uses electrical signals, called action potentials, to activate muscle. The moment-to-moment control of muscle force is graded by the nervous system by varying the rate at which motor units discharge action potentials (rate coding), and by altering the number of motor units that are active (recruitment). A recording of this electrical activity, the electromyogram (EMG), represents the summated activity of both the recruitment and rate coding of those motor units within the detection range of the recording electrodes. Therefore, the EMG signal represents the peripheral neural drive to the muscle, but in a manner that is not clearly understood. Because the interactions of a motor unit population cannot be studied experimentally in humans, and the understanding of these interactions improves only slightly during reduced animal preparations, a computational approach was used to examine the relation between the EMG signal and activation of the motor unit pool. The first study indicated that the surface EMG underestimated the activation signal sent by the nervous system to muscle due to cancellation between the positive and negative phases of motor unit potentials (amplitude cancellation). The second study examined how different motor unit properties influenced the size of potentials evoked with an electrical stimulus. The third study evaluated how extracting motor unit properties by averaging into the EMG signal from the discharges of single neurons (spike-triggered averaging) were limited by amplitude cancellation and the correlated discharge of motor units (motor unit synchronization). The fourth study, which simulated two motor unit pools, determined that although cross-correlation analyses of pairs of EMG signals were influenced by common input to the different motor unit pools, changes in many parameters influenced the degree of correlation, limiting their use as a population measure of motor unit synchronization. These results indicate that although the surface EMG signal is commonly used as an index of activity in populations of motor units, there are limits to the information that can be extracted from this signal. These limitations influence our ability to interpret adaptations within the neuromuscular system from the EMG signal.