Comparison of an accelerometer and a condenser microphone for mechanomyographic signals during measurement of agonist and antagonist muscles in sustained isometric muscle contractions.

Abstract

The purpose of this study was to investigate the influence of the force tremor (FT) on mechanomyographic (MMG) signals recorded by a condenser microphone (MIC) and an accelerometer (ACC) during measurement of agonist and antagonist muscles in sustained isometric contractions. Surface electromyographic (EMG) signals and MMG signals by MIC (MMG-MIC) and ACC (MMG-ACC) were recorded simultaneously on biceps brachii (BB) and triceps brachii (TB). Following determination of the isometric maximum voluntary contraction (MVC), 10 male subjects were asked to perform sustained elbow flexion and extension contractions at 30% MVC until exhaustion. We analyzed the root mean square (RMS) for all signals and compared the sum of the power spectrum (SPA) for 3-6 Hz and 8-12 Hz and the ratio of the sum of SPA for 3-6 Hz and 8-12 Hz in SPA for 3-100 Hz (SPA-FT/SPA-(3-100 Hz)) between MMG-MIC and MMG-ACC. During all sustained muscle contractions, the RMS of EMG and MMG-(MIC) was significantly (p<0.05) increased in antagonistic muscle pairs, while the increase was more noticeable for the agonist than for the antagonist. In addition, the antagonist had a significantly (p<0.05) smaller amplitude than the agonist muscle. The RMS of MMG-ACC, however, showed no significant (p>0.05) difference in RMS amplitude and slope between agonist and antagonist muscles during flexion. In extension, the MMG-ACC-RMS amplitude showed a tendency to be higher in the antagonist than in the agonist, while their slopes showed no significant (p>0.05) difference. The SPA for 3-6 Hz and 8-12 Hz in MMG-(MIC) showed a tendency to be higher in the agonist than the antagonist, and the slopes of the agonist were significantly (p<0.05) higher than those of the antagonist in all contractions. In MMG-ACC, SPA and slopes for 3-6 Hz and 8-12 Hz tended not to differ between agonist and antagonist. The SPA-FT/SPA-(3-100 Hz) in MMG-ACC showed that the antagonist was higher than that of the agonist in all contractions. The MMG-(MIC), however, showed a tendency toward no difference between the agonist and antagonist. In the assessment of muscle activity during simultaneous measurement of the agonist and antagonist during sustained muscle contractions, the MMG signal detected by MIC appeared to be less affected by FT than by ACC due to the different inherent characteristics of the two transducers.