Neural mechanisms of intermuscular coherence : 3 Implications for the rectification of surface electromyography

21 Oscillatory activity plays a crucial role in corticospinal control of muscle synergies 22 and is widely investigated using corticospinal and intermuscular synchronization. 23 However, the neurophysiological mechanisms that translate these rhythmic patterns 24 into surface electromyography (EMG) are not well understood. This is underscored 25 by the ongoing debate on the rectification of surface EMG prior to spectral analysis. 26 Whereas empirical studies commonly rectify surface EMG, computational approaches 27 have argued against it. In the present study we employ a computational model to 28 investigate the role of the motor unit action potential (MAUP) on the translation of 29 oscillatory activity. That is, diverse MUAP shapes may distort the transfer of common 30 input into surface EMG. We test this in a computational model consisting of two 31 motor unit pools receiving common input and compare it to empirical results of 32 intermuscular coherence between bilateral leg muscles. The shape of the MUAP was 33 parametrically varied and power and coherence spectra investigated with and without 34 rectification. The model shows that the effect of EMG rectification depends on the 35 uniformity of MUAP shapes. When output spikes of different motor units are 36 convolved with identical MUAPs, oscillatory input is evident in both rectified and 37 non-rectified EMG. In contrast, a heterogeneous MAUP distribution distorts common 38 input and oscillatory components are only manifest as periodic amplitude 39 modulations, i.e. in rectified EMG. The experimental data showed that intermuscular 40 coherence was mainly discernable in rectified EMG, hence providing empirical 41 support for a heterogeneous distribution of MUAPs. These findings implicate that the 42 shape of MUAPs is an essential parameter to reconcile experimental and 43 computational approaches. 44 45