The research conducted in the last three decades has collectively demonstrated that the skeletal muscle performance can be alternatively assessed by mechanomyographic signal (MMG) parameters. Indices of muscle performance, not limited to force, power, work, endurance and the related physiological processes underlying muscle activities during contraction have been evaluated in the light of the signal features. As a non-stationary signal that reflects several distinctive patterns of muscle actions, the illustrations obtained from the literature support the reliability of MMG in the analysis of muscles under voluntary and stimulus evoked contractions. An appraisal of the standard practice including the measurement theories of the methods used to extract parameters of the signal is vital to the application of the signal during experimental and clinical practices, especially in areas where electromyograms are contraindicated or have limited application. As we highlight the underpinning technical guidelines and domains where each method is well-suited, the limitations of the methods are also presented to position the state of the art in MMG parameters extraction, thus providing the theoretical framework for improvement on the current practices to widen the opportunity for new insights and discoveries. Since the signal modality has not been widely deployed due partly to the limited information extractable from the signals when compared with other classical techniques used to assess muscle performance, this survey is particularly relevant to the projected future of MMG applications in the realm of musculoskeletal assessments and in the real time detection of muscle activity.

The evoked electromyographic signal (eEMG) potential is the standard index used to monitor both electrical changes within the motor unit during muscular activity and the electrical patterns during evoked contraction. However, technical and physiological limitations often preclude the acquisition and analysis of the signal especially during functional electrical stimulation (FES)-evoked contractions. Hence, an accurate quantification of the relationship between the eEMG potential and FES-evoked muscle response remains elusive and continues to attract the attention of researchers due to its potential application in the fields of biomechanics, muscle physiology, and rehabilitation science. We conducted a systematic review to examine the effectiveness of eEMG potentials to assess muscle force and fatigue, particularly as a biofeedback descriptor of FES-evoked contractions in individuals with spinal cord injury. At the outset, 2867 citations were identified and, finally, fifty-nine trials met the inclusion criteria. Four hypotheses were proposed and evaluated to inform this review. The results showed that eEMG is effective at quantifying muscle force and fatigue during isometric contraction, but may not be effective during dynamic contractions including cycling and stepping. Positive correlation of up to r = 0.90 (p < 0.05) between the decline in the peak-to-peak amplitude of the eEMG and the decline in the force output during fatiguing isometric contractions has been reported. In the available prediction models, the performance index of the eEMG signal to estimate the generated muscle force ranged from 3.8% to 34% for 18 s to 70 s ahead of the actual muscle force generation. The strength and inherent limitations of the eEMG signal to assess muscle force and fatigue were evident from our findings with implications in clinical management of spinal cord injury (SCI) population.

The difficulty of real-time muscle force or joint torque estimation during neuromuscular electrical stimulation (NMES) in physical therapy and exercise science has motivated recent research interest in torque estimation from other muscle characteristics. This study investigated the accuracy of a computational intelligence technique for estimating NMES-evoked knee extension torque based on the Mechanomyographic signals (MMG) of contracting muscles that were recorded from eight healthy males. Simulation of the knee torque was modelled via Support Vector Regression (SVR) due to its good generalization ability in related fields. Inputs to the proposed model were MMG amplitude characteristics, the level of electrical stimulation or contraction intensity, and knee angle. Gaussian kernel function, as well as its optimal parameters were identified with the best performance measure and were applied as the SVR kernel function to build an effective knee torque estimation model. To train and test the model, the data were partitioned into training (70%) and testing (30%) subsets, respectively. The SVR estimation accuracy, based on the coefficient of determination (R²) between the actual and the estimated torque values was up to 94% and 89% during the training and testing cases, with root mean square errors (RMSE) of 9.48 and 12.95, respectively. The knee torque estimations obtained using SVR modelling agreed well with the experimental data from an isokinetic dynamometer. These findings support the realization of a closed-loop NMES system for functional tasks using MMG as the feedback signal source and an SVR algorithm for joint torque estimation.

A mechanomyography muscle contraction (MC) sensor, affixed to the skin surface, was used to quantify muscle tension during repetitive functional electrical stimulation (FES)-evoked isometric rectus femoris contractions to fatigue in individuals with spinal cord injury (SCI). Nine persons with motor complete SCI were seated on a commercial muscle dynamometer that quantified peak torque and average torque outputs, while measurements from the MC sensor were simultaneously recorded. MC-sensor-predicted measures of dynamometer torques, including the signal peak (SP) and signal average (SA), were highly associated with isometric knee extension peak torque (SP: r = 0.91, p < 0.0001), and average torque (SA: r = 0.89, p < 0.0001), respectively. Bland-Altman (BA) analyses with Lin’s concordance (ρ_C) revealed good association between MC-sensor-predicted peak muscle torques (SP; ρ_C = 0.91) and average muscle torques (SA; ρ_C = 0.89) with the equivalent dynamometer measures, over a range of FES current amplitudes. The relationship of dynamometer torques and predicted MC torques during repetitive FES-evoked muscle contraction to fatigue were moderately associated (SP: r = 0.80, p < 0.0001; SA: r = 0.77; p < 0.0001), with BA associations between the two devices fair-moderate (SP; ρ_C = 0.70: SA; ρ_C = 0.30). These findings demonstrated that a skin-surface muscle mechanomyography sensor was an accurate proxy for electrically-evoked muscle contraction torques when directly measured during isometric dynamometry in individuals with SCI. The novel application of the MC sensor during FES-evoked muscle contractions suggested its possible application for real-world tasks (e.g., prolonged sit-to-stand, stepping,) where muscle forces during fatiguing activities cannot be directly measured.