Research On Static Muscle Fatigue Trend Via Nonlinear Feature Analysis Of Mechanomyography Signal

Muscle fatigue is the embodiment of the decrease of the ability of muscles,which has been applied in human body exercise,rehabilitation training and occupational disease prevention.Muscle fatigue is a phenomenon in which the maximum voluntary muscle contraction force decreases during the process of muscle contractions.Mechanomyography(MMG)signal is a nonlinear signal that records the mechanical vibration characteristics of’muscle fibers during contraction.By analyzing the MMG signal of the upper trapezius muscle when standing and shrugging the shoulders statically,the relationship between the nonlinear dynamic feature of MMG signal and muscle fatigue were found.In this study,MMG of the upper trapezius muscle were collected from 10 healthy subj ects when they performed static shoulder shrugging with a constant force of 50%maximum voluntary contractile force(MVC),and noise reduction,segmentation and feature extraction were performed on MMG signal to analyze the corresponding relationship between the trend of feature values and muscle fatigue.In process of signal noise reduction,three empirical mode decomposition(EMD)algorithms were compared,and a complete ensemble EMD with adaptive noise combined correlation coefficient was adopted.In features analysis,the features of periodic signal,MMG signal and random signal in fractal dimension(FD),LZ complexity(LZC),maximum Lyapunov exponent(MLE)and fuzzy entropy(FEN)were compared.In feature extraction,four nonlinear dynamic features were used to analyze the fatigue state of the upper trapezius muscle during static contraction and compared with the time-domain features such as median frequency(MDF)and average power frequency(MPF).The results showed that the noise reduction method based on CEEMDAN algorithm combined correlation coefficient was effective.MMG signal is a nonlinear and non-stationary signal with chaotic characteristics.The nonlinear characteristics of MMG signal are between periodic signal and random signal.FD,LZC,MLE and FEN of MMG signal showed an approximately linear decline with the deepening of muscle fatigue,which was similar to the effect of MDF and MPF in muscle fatigue assessment.Therefore,as a quantification of the fatigue state of muscles,the features of MMG signals would be well marked and measured by FD,LZC,MLE,FEN and other nonlinear features which have a certain advantage in terms of applicability and a certain degree of theoretical support in the sensitivity,in addition to the traditional MDF and MPF,providing a new method for MMG signal to evaluate muscle fatigue in nonlinear field.