Upper Arm Motion High-Density sEMG Recognition Optimization Based on Spatial and Time-Frequency Domain Features

Abstract

Background. Spatial characteristics of sEMG signals are obtained by high-density matrix sEMG electrodes for further complex upper arm movement classification. Multiple electrode channels of the high-density sEMG acquisition device aggravate the burden of the microprocessor and deteriorate control system’s real-time performance at the same time. A shoulder motion recognition optimization method based on the maximizing mutual information from multiclass CSP selected spatial feature channels and wavelet packet features extraction is proposed in this study. Results. The relationship between the number of channels and recognition rate is obtained by the recognition optimization method. The original 64 electrodes channels are reduced to only 4-5 active signal channels with the accuracy over 92%. Conclusion. The shoulder motion recognition optimization method is combined with the spatial-domain and time-frequency-domain features. In addition, the spatial feature channel selection is independent of feature extraction and classification algorithm. Therefore, it is more convenient to use less channels to achieve the desired classification accuracy.

Lw-CNN-Based Myoelectric Signal Recognition and Real-Time Control of Robotic Arm for Upper-Limb Rehabilitation

“…It is very difficult to recognize body motion intents in the context of coordinated muscle effect [ 19 ]. Additional sEMG or HD-sEMG channels and more complicated classification models are often used to increase classification accuracy in studies concerning gesture recognition [ 49 ]. However, this greatly increases the amount of system computation necessary making these methods unsuitable for real-time control.…”

Spatial Filtering for Enhanced High-Density Surface Electromyographic Examination of Neuromuscular Changes and Its Application to Spinal Cord Injury

“…The supervised methods are trained to learn some generalizable regulations in transformation of the HD-sEMG data into a set of components driven by well-labeled data. For example, common spatial pattern (CSP) is a supervised method that searches the generalized eigenvector or the projection vector to maximize the variance between different predefined classes so as to facilitate classification of HD-sEMG patterns [25], [26]. The other category always employs unsupervised matrix factorization algorithms to process multi-channel sEMG and HD-sEMG data.…”

Spatial Filtering for Enhanced High-Density Surface Electromyographic Examination of Neuromuscular Changes and Its Application to Spinal Cord Injury

“…The spatial filtering technique can be employed to remove artifacts of HD-sEMG data and to retain useful information given the muscular activation heterogeneity. Its basic principle is to preserve the sources of interest and suppress unwanted components from signals [21][22][23][24][25][26]. Various matrix factorization algorithms [15], [16], [27][28][29][30][31][32][33][34][35] relied on different criteria concerning inherent structure of the input multivariate data.…”