sEMG-Based Neural Network Prediction Model Selection of Gesture Fatigue and Dataset Optimization

Abstract

The fatigue energy consumption of independent gestures can be obtained by calculating the power spectrum of surface electromyography (sEMG) signals. The existing research studies focus on the fatigue of independent gestures, while the research studies on integrated gestures are few. However, the actual gesture operation mode is usually integrated by multiple independent gestures, so the fatigue degree of integrated gestures can be predicted by training neural network of independent gestures. Three natural gestures including browsing information, playing games, and typing are divided into nine independent gestures in this paper, and the predicted model is established and trained by calculating the energy consumption of independent gestures. The artificial neural networks (ANNs) including backpropagation (BP) neural network, recurrent neural network (RNN), and long short-term memory (LSTM) are used to predict the fatigue of gesture. The support vector machine (SVM) is used to assist verification. Mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) are utilized to evaluate the optimal prediction model. Furthermore, the different datasets of the processed sEMG signal and its decomposed wavelet coefficients are trained, respectively, and the changes of error functions of them are compared. The experimental results show that LSTM model is more suitable for gesture fatigue prediction. The processed sEMG signals are appropriate for using as the training set the fatigue degree of one-handed gesture. It is better to use wavelet decomposition coefficients as datasets to predict the high-dimensional sEMG signals of two-handed gestures. The experimental results can be applied to predict the fatigue degree of complex human-machine interactive gestures, help to avoid unreasonable gestures, and improve the user’s interactive experience.

Retentive capacity of power output and linear versus non-linear mapping of power loss in the isotonic muscular endurance test

“…However, another study showed that the linear and non-linear approaches were equally valid to estimate changes in power loss during a fatiguing repetitive leg extension exercise 12 , 31 . In the present study, non-linear techniques were developed using an MLPNN to combine different time and spectral features of the sEMG signal into a fatigue index representing the estimated changes in power output 12 , 31 , 34 , 35 . The results showed that the estimation errors were smaller when using non-linear mapping of power loss during dynamic contractions compared to linear mapping, as it showed higher SNR and correlation coefficients between the actual and estimated power output in three muscle groups of lower limbs.…”

“…A multi-layer perception neural network (MLPNN) was chosen to relate changes in sEMG variables and power output because it shows good accuracy to relate sEMG variables and fatigue indices 12 , 31 , 35 . Four sEMG parameters were calculated from each contraction of all of the subjects: (1) MAV is an estimate of the mean absolute value of the signal, as the integrated EMG is divided by the integration time 15 , 34 , 50 ; (2) zero crossing (ZC), as the number of times that the waveform crosses zero, is a simple measure of the main frequency of the signal 15 , 34 , 50 ; (3) slope sign change (SSC), as the number of times that the slope of the waveform changes sign, provides another measure of frequency content 15 , 34 , 50 ; and (4) wavelength (WL) provides information on the waveform complexity in each segment 15 , 34 , 50 .…”

“…Moreover, the relationship between sEMG and power loss is useful for assessing and understanding neuromuscular fatigue, as it can be defined on the basis of electrophysiological or mechanical events 30 – 33 , and linear techniques are often used to estimate muscle fatigue by relating changes in sEMG parameters to changes in power loss. However, myoelectric signals can be better modeled as outputs of a non-linear dynamic system rather than as random stochastic signals, so a non-linear model based on a learning procedure would provide more accurate tracking of power loss using sEMG variables compared to a linear model 12 , 31 , 34 , 35 . Additionally, combining the subjective estimates of load tolerance and the subjects’ determinations of their own muscular endurance is beneficial for revealing the subjects’ real-world capacity during muscular endurance experiments 20 , 36 , and the results can help balance the workload and physical capacity to prevent acute or overuse injuries in actual application 37 , 38 .…”

Implementation of surface electromyography controlled prosthetics limb based on recurrent neural network

“…They are a special type of RNN that attempts to solve the vanishing gradient problem and is widely used in sEMG control like performing models for myoelectric based control of a prosthetic hand, [8][9][10] neural interface toward intuitive prosthetic control for amputees, 11 control of a hybrid dynamical transfemoral prosthesis, 12 prediction model of gesture, and dataset optimization. [13][14][15] Deep learning techniques involve large neural networks and achieve state-of-the-art output pattern recognition accuracy. This is done in two stages: in the learning phase, a classifier is created based on the input signal (sEMG) and output value to determine which signal belongs to.…”

“…To address this issue, long‐short term memory (LSTM) networks have been developed. They are a special type of RNN that attempts to solve the vanishing gradient problem and is widely used in sEMG control like performing models for myoelectric based control of a prosthetic hand, 8–10 neural interface toward intuitive prosthetic control for amputees, 11 control of a hybrid dynamical transfemoral prosthesis, 12 prediction model of gesture, and dataset optimization 13–15 …”