Spectral moments of mechanomyographic signals recorded with accelerometer and microphone during sustained fatiguing contractions.

The aim of this study was to analyse the trends of the first three power spectral moments of the mechanomyogram (MMG) signal recorded by a microphone (MMG(MIC)) and an accelerometer (MMG(ACC)) during sustained contractions. MMG signals were recorded from the biceps brachii muscle in 14 healthy male subjects during a 3 min isometric elbow flexion at 30% of the maximal voluntary contraction. MMG absolute and normalised root mean square (RMS), mean power frequency (MNF), power spectral variance (Mc2), and skewness (mu3) were computed. For both MMG(MIC) and MMG(ACC), absolute and normalised RMS and Mc2 increased while MNF and mu3 decreased with contraction time (P<0.001). The rates of change of RMS over time were significantly correlated (P<0.001) for MMG(MIC) and MMG(ACC) but not correlated for spectral moments. The coefficient of variation of RMS was higher for MMG(MIC) than for MMG(ACC), while the opposite was observed for mu3 (P<0.05). It was concluded that higher order spectral moments of the MMG signal change during sustained contraction, indicating a complex modification of the shape of the power spectrum and not just scaling of the bandwidth. This is most likely due to the additional motor unit recruitment with fatigue and to the non-linear summation of motor unit contributions to the signal. Moreover, the characteristics of MMG signals recorded with microphones and accelerometers have important differences, which should be taken into account when comparing results from different studies.