Accurate measurement of inhibition and excitation in human motor pools during sensory stimulation

Surface electromyography (sEMG) is a pivotal approach in clinical and basic neurophysiology, allowing us to extract the summed activity of motor units in a given muscle. Due to the bipolar nature of the motor unit action potential, the sEMG is a non-linear representation of their underlying motor unit activity and therefore affected by signal cancellation. It is not clear how this cancellation influences evoked responses in sEMG. The aim of our study was to characterise how representative an evoked sEMG response was to the firing behaviour of the underlying motor pool. To do this, we first simulated a population of motor units (and their action potentials) that responded to a stimulus with a change in firing probability. Their activity was summed then rectified to generate a simulated sEMG signal or was rectified and then summed, to generate a sEMG signal with no cancellation. By comparing the two responses to that of the underlying pool we would then compare for discrepancies. We repeated this process but by using the responses of tibialis anterior motor units to weak tibial nerve stimulation. We find that both for the simulated and experimental data the response measured through the sEMG is almost always an underestimate of the evoked response in the underlying motor pool. This is the case for both inhibitory and excitatory evoked responses. The magnitude of the inaccuracy depends on the size of the evoked response, but it cannot be accounted solely by signal cancellation, suggesting other factors may also contribute.