The automatic recognition technology of muscle fatigue has widespread application in the field of kinesiology and rehabilitation medicine. In this paper, we used surface electromyography (sEMG) to study the recognition of leg muscle fatigue during circuit resistance training. The purpose of this study was to solve the problem that the sEMG signals have a lot of noise interference and the recognition accuracy of the existing muscle fatigue recognition model is not high enough. First, we proposed an improved wavelet threshold function denoising algorithm to denoise the sEMG signal. Then, we build a muscle fatigue state recognition model based on long short-term memory (LSTM), and used the Holdout method to evaluate the performance of the model. Finally, the denoising effect of the improved wavelet threshold function denoising method proposed in this paper was compared with the denoising effect of the traditional wavelet threshold denoising method. We compared the performance of the proposed muscle fatigue recognition model with that of particle swarm optimization support vector machine (PSO-SVM) and convolutional neural network (CNN). The results showed that the new wavelet threshold function had better denoising performance than hard and soft threshold functions. The accuracy of LSTM network model in identifying muscle fatigue was 4.89% and 2.47% higher than that of PSO-SVM and CNN, respectively. The sEMG signal denoising method and muscle fatigue recognition model proposed in this paper have important implications for monitoring muscle fatigue during rehabilitation training and exercise.
Surface electromyography (sEMG) has been widely used in the study of clinical medicine, rehabilitation medicine, sports, etc., and its endpoints should be detected accurately before analyzing. However, endpoint detection is vulnerable to electrocardiogram (ECG) interference when the sEMG recorders are placed near the heart. In this paper, an endpoint-detection algorithm which is insensitive to ECG interference is proposed. In the algorithm, endpoints of sEMG are detected based on the short-time energy and short-time zero-crossing rates of sEMG. The thresholds of short-time energy and short-time zero-crossing rate are set according to the statistical difference of short-time zero-crossing rate between sEMG and ECG, and the statistical difference of short-time energy between sEMG and the background noise. Experiment results on the sEMG of rectus abdominis muscle demonstrate that the algorithm detects the endpoints of the sEMG with a high accuracy rate of 95.6%.
In this paper, a new surface electromyography (sEMG) signal decomposition method based on spatial location is proposed for the high-density sEMG signals in dynamic muscle contraction. Firstly, according to the waveform correlation of each muscle motor units (MU) in each channel, the firing times are extracted, and then the firing times are classified by the spatial location of MU. The MU firing trains are finally obtained. The simulation results show that the accuracy rate of a single MU firing train after classification is more than 91.67%. For real sEMG signals, the accuracy rate to find a same MU by the “two source” method is over (88.3 ± 2.1)%. This paper provides a new idea for dynamic sEMG signal decomposition.
In this study, surface electromyography (sEMG) of the lower limbs of cerebral-palsy (CP) subjects in gait cycle was recorded and its parameters of gait cycle characters were analyzed to assess their clinical severity. Three algorithms, including integrated profile (IP), sample-entropy (SampEN) and smooth nonlinear energy operator (SNEO) algorithm, were applied to calculate the duration of walking sEMG segments in simulated SEMG signals. After that, the efficiency and accuracy were compared among these three algorithms. SNEO was then selected as the optimal algorithm among the three algorithms and employed for real sEMG signal processing of CP subjects. The results indicated that there was no significant difference in the accuracy of sEMG segement detection for the three algorithms. However, the computation speed of SNEO algorithm was much faster than those of the others and thus it was a suitable algorithm for detecting walking sEMG segments of CP subjects. In addition, the positive correlation was found between the clinical severity and the mean duration of walking sEMG segments in CP subjects. The results indicated that there was a significant difference in the three groups of CP subjects with different levels of severity. Our findings showed that the mean duration of walking sEMG segments could be considered as an assistant index to evaluate the clinical severity of CP subjects.
Exercise-induced muscle fatigue is a phenomenon that the maximum voluntary contraction force or power output of muscle is temporarily reduced due to muscular movement. If the fatigue is not treated properly, it will bring about a severe injury to the human body. With multi-channel collection of lower limb surface electromyography signals, this article analyzes the muscle fatigue by adoption of band spectrum entropy method which combined electromyographic signal spectral analysis and nonlinear dynamics. The experimental result indicated that with the increase of muscle fatigue, muscle signal spectrum began to move to low frequency, the energy concentrated, the system complexity came down, and the band spectrum entropy which reflected the complexity was also reduced. By monitoring the entropy, we can measure the degree of muscle fatigue, and provide an indicator to judge fatigue degree for the sports training and clinical rehabilitation training.
To better analyze the problem of abnormal neuromuscular coupling related to motor dysfunction for stroke patients, the functional coupling of the multichannel electromyography (EMG) were studied and the difference between stroke patients and healthy subjects were further analyzed to explore the pathological mechanism of motor dysfunction after stroke. Firstly, the cross-frequency coherence (CFC) analysis and non-negative matrix factorization (NMF) were combined to construct a CFC-NMF model to study the linear coupling relationship in bands and the nonlinear coupling characteristics in different frequency ratios during elbow flexion and extension movement. Furthermore, the significant coherent area and sum of cross-frequency coherence were respectively calculated to quantitatively describe the intermuscular linear and nonlinear coupling characteristics. The results showed that the linear coupling relationship between multichannel muscles was different in frequency bands and the overall coupling was stronger in low frequency band. The linear coupling strength of the stroke patients was lower than that of the healthy subjects in different frequency bands especially in beta and gamma bands. For the nonlinear coupling, the intermuscular coupling strength of stroke patients in different frequency ratios was significantly lower than that of the healthy subjects, and the coupling strength in the frequency ratio 1∶2 was higher than that in the frequency ratio 1∶3. This method can provide a theoretical basis for exploring the intermuscular coupling mechanism of patients with motor dysfunction.
The real physical image of the affected limb, which is difficult to move in the traditional mirror training, can be realized easily by the rehabilitation robots. During this training, the affected limb is often in a passive state. However, with the gradual recovery of the movement ability, active mirror training becomes a better choice. Consequently, this paper took the self-developed shoulder joint rehabilitation robot with an adjustable structure as an experimental platform, and proposed a mirror training system completed by next four parts. First, the motion trajectory of the healthy limb was obtained by the Inertial Measurement Units (IMU). Then the variable universe fuzzy adaptive proportion differentiation (PD) control was adopted for inner loop, meanwhile, the muscle strength of the affected limb was estimated by the surface electromyography (sEMG). The compensation force for an assisted limb of outer loop was calculated. According to the experimental results, the control system can provide real-time assistance compensation according to the recovery of the affected limb, fully exert the training initiative of the affected limb, and make the affected limb achieve better rehabilitation training effect.
Based on the structure and motion bionic principle of the normal adult fingers, biological characteristics of human hands were analyzed, and a wearable exoskeleton hand function training device for the rehabilitation of stroke patients or patients with hand trauma was designed. This device includes the exoskeleton mechanical structure and the electromyography (EMG) control system. With adjustable mechanism, the device was capable to fit different finger lengths, and by capturing the EMG of the users’ contralateral limb, the motion state of the exoskeleton hand was controlled. Then driven by the device， the user’s fingers conducting adduction/abduction rehabilitation training was carried out. Finally, the mechanical properties and training effect of the exoskeleton hand were verified through mechanism simulation and the experiments on the experimental prototype of the wearable exoskeleton hand function training device.
The present study was carried out with the surface electromyography signal of subjects during the time when subjects did the exercises of the 6 core stability trainings. We analyzed the different activity level of surface electromyography signal, and finally got various fatigue states of muscles in different exercises. Thirty subjects completed exercises of 6 core stability trainings, which were prone bridge, supine bridge, unilateral bridge (divided into two trainings,i.e. the left and right sides alternatively) and bird-dog (divided into two trainings,i.e. the left and right sides alternatively), respectively. Each exercise was held on for 1 minute and 2 minutes were given to relax between two exercises in this test. We measured both left and right sides of the body’s muscles, which included erector spina, external oblique, rectus abdominis, rectus femoris, biceps femoris, anterior tibial and gastrocnemius muscles. We adopted the frequency domain characteristic value of the surface electromyography signal,i.e. median frequency slope to analyze the muscle fatigue in this study. In the present paper, the results exhibit different fatigue degrees of the above muscles during the time when they did the core stability rehabilitation exercises. It could be concluded that supine bridge and unilateral bridge can cause more fatigue on erector spina muscle, prone bridge caused Gastrocnemius muscle much fatigue and there were statistical significant differences (P<0.05) between prone bridge and other five rehabilitation exercises in the degree of rectus abdominis muscle fatigue. There were no statistical significant differences (P>0.05) between all the left and right sides of the same-named muscles in the median frequency slope during all the exercises of the six core stability trainings,i.e. the degree which the various kinds of rehabilitation exercises effected the left and right side of the same-named muscle had no statistical significant difference (P>0.05). In this research, the conclusion presents quantized guidelines on the effects of core stability trainings on different muscles.
At present, upper limb motor rehabilitation relies on specific rehabilitation aids, ignoring the initiative of upper limb motor of patients in the middle and late stages of rehabilitation. This paper proposes a fuzzy evaluation method for active participation based on trajectory error and surface electromyography (sEMG) for patients who gradually have the ability to generate active force. First, the level of motor participation was evaluated using trajectory error signals represented by computer vision. Then, the level of physiological participation was quantified based on muscle activation (MA) characterized by sEMG. Finally, the motor performance and physiological response parameters were input into the fuzzy inference system (FIS). This system was then used to construct the fuzzy decision tree (FDT), which ultimately outputs the active participation level. A controlled experiment of upper limb flexion and extension exercise in 16 healthy subjects demonstrated that the method presented in this paper was effective in quantifying difference in the active participation level of the upper limb in different force-generating states. The calculation results of this method and the active participation assessment method based on sEMG during the task cycle showed that the active participation evaluation values of both methods peaked in the initial cycle: (82.34 ± 9.3) % for this paper’s method and (78.44 ± 7.31) % for the sEMG method. In the subsequent cycles, the values of both showed a dynamic change trend of rising first and then falling. Trend consistency verifies the effectiveness of the active participation assessment strategy in this paper, providing a new idea for quantifying the participation level of patients in middle and late stages of upper limb rehabilitation without special equipment mediation.