Continuous online prediction of lower limb joints angles based on sEMG signals by deep learning approach

• Published in: Computers in biology and medicine Vol. 163; p. 107124
• Main Authors: Song, Qiuzhi, Ma, Xunju, Liu, Yali
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.09.2023; Elsevier Limited
• Subjects: Accuracy; Algorithms; Cooperative control; Correlation coefficients; Deep learning; Electromyography; Feature extraction; Gait; Human mechanics; Internal Medicine; Internet; Knee; Long short-term memory; LSTM; Machine learning; Muscles; Neural networks; Online joint angle prediction; Other; Performance enhancement; Phases; Plantar pressure; Prediction models; Prostheses; Robots; sEMG; Sensors; Standardization; sEMG; LSTM; Online joint angle prediction; Machine learning
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2023.107124

Continuous online prediction of human joints angles is a key point to improve the performance of man-machine cooperative control. In this study, a framework of online prediction method of joints angles by long short-term memory (LSTM) neural network only based on surface electromyography (sEMG) signals was proposed. The sEMG signals from eight muscles of five subjects’ right leg and three joints angles and plantar pressure signals of subjects were collected simultaneously. Different inputs (only sEMG (unimodal), sEMG combined with plantar pressure (multimodal)) after online feature extraction and standardization were used for training the angle online prediction model by LSTM. The results indicate that there is no significant difference between the two kinds of inputs for LSTM model and the proposed method can make up for the shortage of using a single type of sensor. The range of mean values of root square mean error, mean absolute error and Pearson correlation coefficient of the three joints angles achieved by the proposed model only with the input of sEMG under four kinds of predicted time (50, 100, 150, and 200 ms) are [1.63°,3.20°],[1.27°, 2.36°] and [0.9747, 0.9935]. Three popular machine learning algorithms with different inputs were compared to the proposed model only based on sEMG. Experiment results demonstrate that the proposed method has the best prediction performance and there are highly significant differences between it and other methods. The difference of prediction results under different gait phases by the proposed method was also analyzed. The results indicate that the prediction effect of support phases is generally better than that of swing phases. Above experimental results show that the proposed method can realize accurate online joint angle prediction and has better performance to promote man-machine cooperation. [Display omitted] •This study proposed a framework of sEMG data online preprocessing to acquire the root mean square (RMS) feature of sEMG signals with standardization online. And a construction mothed of the training and test set for training the prediction model was proposed. On this basis, a framework of online joint angle prediction of lower limb was proposed.•LSTM neural network was used for predicting the lower limb joints angles, and high prediction accuracy under four kinds of prediction time was achieved.•The effects of different inputs (sEMG (unimodal signals), sEMG combined with plantar pressure (multimodal signals)) on the results of model were compared and analyzed. And the results demonstrate there is no significant difference between the two inputs for the proposed model.•Compared with other three machine learning algorithms with different inputs, the significant superiority of the proposed method only based on sEMG was demonstrated.•Difference of prediction results under different gait phases by the proposed method was analyzed. An important conclusion was found. It is that the prediction effect of support phases is generally better than that of swing phases.