Adaptive Spatial Filtering of High-Density EMG for Reducing the Influence of Noise and Artefacts in Myoelectric Control

Exploring the EMG transient: the muscular activation sequences used as novel time-domain features for hand gestures classification

Introduction

Muscular activation sequences have been shown to be suitable time-domain features for classification of motion gestures. However, their clinical application in myoelectric prosthesis control was never investigated so far. The aim of the paper is to evaluate the robustness of these features extracted from the EMG signal in transient state, on the forearm, for classifying common hand tasks.

Methods

The signal associated to four hand gestures and the rest condition were acquired from ten healthy people and two persons with trans-radial amputation. A feature extraction algorithm allowed for encoding the EMG signals into muscular activation sequences, which were used to train four commonly used classifiers, namely Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), Non-linear Logistic Regression (NLR) and Artificial Neural Network (ANN). The offline performances were assessed with the entire sample of recruited people. The online performances were assessed with the amputee subjects. Moreover, a comparison of the proposed method with approaches based on the signal envelope in the transient state and in the steady state was conducted.

Results

The highest performance were obtained with the NLR classifier. Using the sequences, the offline classification accuracy was higher than 93% for healthy and amputee subjects and always higher than the approach with the signal envelope in transient state. As regards the comparison with the steady state, the performances obtained with the proposed method are slightly lower (<4%), but the classification occurred at least 200 ms earlier. In the online application, the motion completion rate reached up to 85% of the total classification attempts, with a motion selection time that never exceeded 218 ms.

Discussion

Muscular activation sequences are suitable alternatives to the time-domain features commonly used in classification problems belonging to the sole EMG transient state and could be potentially exploited in control strategies of myoelectric prosthesis hands.

Movement recognition via channel-activation-wise sEMG attention

CONTEXT

Surface electromyography (sEMG) signals contain rich information recorded from muscle movements and therefore reflect the user's intention. sEMG has seen dominant applications in rehabilitation, clinical diagnosis as well as human engineering, etc. However, current feature extraction methods for sEMG signals have been seriously limited by their stochasticity, transiency, and non-stationarity.

OBJECTIVE

Our objective is to combat the difficulties induced by the aforementioned downsides of sEMG and thereby extract representative features for various downstream movement recognition.

METHOD

We propose a novel 3-axis view of sEMG features composed of temporal, spatial, and channel-wise summary. We leverage the state-of-the-art architecture Transformer to enforce efficient parallel search and to get rid of limitations imposed by previous work in gesture classification. The transformer model is designed on top of an attention-based module, which allows for the extraction of global contextual relevance among channels and the use of this relevance for sEMG recognition.

RESULTS

We compared the proposed method against existing methods on two Ninapro datasets consisting of data from both healthy people and amputees. Experimental results show the proposed method attains the state-of-the-art (SOTA) accuracy on both datasets. We further show that the proposed method enjoys strong generalization ability: a new SOTA is achieved by pretraining the model on a different dataset followed by fine-tuning it on the target dataset.

Electroencephalography Signal Processing: A Comprehensive Review and Analysis of Methods and Techniques

The electroencephalography (EEG) signal is a noninvasive and complex signal that has numerous applications in biomedical fields, including sleep and the brain-computer interface. Given its complexity, researchers have proposed several advanced preprocessing and feature extraction methods to analyze EEG signals. In this study, we analyze a comprehensive review of numerous articles related to EEG signal processing. We searched the major scientific and engineering databases and summarized the results of our findings. Our survey encompassed the entire process of EEG signal processing, from acquisition and pretreatment (denoising) to feature extraction, classification, and application. We present a detailed discussion and comparison of various methods and techniques used for EEG signal processing. Additionally, we identify the current limitations of these techniques and analyze their future development trends. We conclude by offering some suggestions for future research in the field of EEG signal processing.

Random Channel Masks for Regularization of Least Squares-Based Finger EMG-Force Modeling to Improve Cross-Day Performance

Estimating the finger forces from surface electromyography (sEMG) is essential for diverse applications (e.g., human-machine interfacing). The performance of pre-trained sEMG-force models degenerates significantly when applied on a second day, due to the large cross-day variation of sEMG characteristics. Previous studies mainly employed transfer learning algorithms to tackle this problem. However, transfer learning algorithms normally require data collected on the second day for model calibration, increasing the inconvenience in practical use. In this work, we investigated the effect of model regularization on this issue. Specifically, 256-channel high-density sEMG (HDsEMG) signals with varying finger forces were collected on different days (3-25 days apart). We applied randomly generated channel perturbations ("masks") to feature maps of randomly selected channels in training dataset. The channel masks of the training set were generated randomly and independently in each narrow time window (~20 ms). We assumed that by learning from randomly masked feature maps (randomness is the central aspect), the model would not be biased by a small number of features but would be based on learning from a global perspective, therefore avoiding overfitting to the within-day EMG patterns. Moore-Penrose inverse model regularization was also employed as a baseline method, with results showing that cross-day EMG-force models require a higher tolerance parameter compared with within-day applications. In combination with the Moore-Penrose inverse model regularization, further applying random channel masks to the training set significantly improved model performance in cross-day validation.

Classification and Regression of Muscle Neural Signals on Human Lower Extremities via BP_AdaBoost

Electromyography (EMG) signals are widely applied in the classification of human motion and intention recognition as having the characteristic of earlier than actual limb motion. In this article, to improve its accuracy of classification and prediction, we firstly analyze the relationship between muscle length and joint movement and select rectus femoris and biceps femoris as the experimental muscles to collect neural signals by means of musculoskeletal analysis software. EMG sensors are used to measure those muscles’ EMG signals of five kinds of knee movements, including thigh-raising, calf-raising, squatting, knee bending on chair, and walking. We designed a BP_AdaBoost algorithm with the BP neural network as a weak classifier and weak regressor, and a muscle neural activation is used as the input for recognition. It is a negative correlation between the length of the rectus femoris and the biceps femoris during gait. Their muscle neural signals are used as the input of the recognition algorithm. The experiment results show that the proposed algorithm improves the rate of BP neural network from 78.82% to 93.52%. The thigh EMG signal successfully maps the knee joint angle by utilizing BP_AdaBoost; its error in identifying five kinds of motion modes is lowest compared with other regression algorithms.

Spatio-temporal warping for myoelectric control: an offline, feasibility study

Objective.The efficacy of an adopted feature extraction method directly affects the classification of the electromyographic (EMG) signals in myoelectric control applications. Most methods attempt to extract the dynamics of the multi-channel EMG signals in the time domain and on a channel-by-channel, or at best pairs of channels, basis. However, considering multi-channel information to build a similarity matrix has not been taken into account.Approach.Combining methods of long and short-term memory (LSTM) and dynamic temporal warping, we developed a new feature, called spatio-temporal warping (STW), for myoelectric signals. This method captures the spatio-temporal relationships of multi-channels EMG signals.Main results. Across four online databases, we show that in terms of average classification error and standard deviation values, the STW feature outperforms traditional features by 5%-17%. In comparison to the more recent deep learning models, e.g. convolutional neural networks (CNNs), STW outperformed by 5%-18%. Also, STW showed enhanced performance when compared to the CNN + LSTM model by 2%-14%. All differences were statistically significant with a large effect size.Significance.This feasibility study provides evidence supporting the hypothesis that the STW feature of the EMG signals can enhance the classification accuracy in an explainable way when compared to recent deep learning methods. Future work includes real-time implementation of the method and testing for prosthesis control.

Combined Dynamic Time Warping and Spatiotemporal Attention for Myoelectric Control

The success of pattern recognition based upper-limb prostheses control is linked to their ability to extract appropriate features from the electromyogram (EMG) signals. Traditional EMG feature extraction (FE) algorithms fail to extract spatial and inter-temporal information from the raw data, as they consider the EMG channels individually across a set of sliding windows with some degree of overlapping. To tackle these limitations, this paper presents a method that considers the spatial information of multi-channel EMG signals by utilising dynamic time warping (DTW). To satisfy temporal considerations, inspired by Long Short-Term Memory (LSTM) neural networks, our algorithm evolves the DTW feature representation across long and short-term components to capture the temporal dynamics of the EMG signal. As such the contribution of this paper is the development of a recursive spatio-temporal FE method, denoted as Recursive Temporal Warping (RTW). To investigate the performance of the proposed method, an offline EMG pattern recognition study with 53 movement classes performed by 10 subjects wearing 8 to 16 EMG channels was considered with the results compared against several conventional as well as deep learning-based models. We show that the use of the RTW can reduce classification errors significantly, paving the way for future real-time implementation.

Contribution of the peroneus longus neuromuscular compartments to eversion and plantarflexion of the ankle

Compartmentalization of animal and human skeletal muscle by multiple motor nerve branches known as the neuromuscular compartment (NMC) has been observed primarily in muscles that participate in a plane of motion. In this context, the peroneus longus muscle contributes to eversion and plantarflexion of the ankle and the presence of NMCs has been reported. However, no research has reported the selective activation of the compartments of the peroneus longus during the performance of different ankle movements. The purpose of this research was to determine the contribution of peroneus longus NMCs, through multi-channel surface electromyography (sEMG), to eversion and plantarflexion movements. Multi-channel sEMG was recorded from the peroneus longus muscle by using an electrode grid during eversion and plantarflexion of the ankle at 10%, 30%, 50%, and 70% of maximal voluntary isometric contraction (MVIC). The root mean square and displacement of the center of mass position in the X (COMx) and Y (COMy) components were calculated. The primary finding was that eversion showed significantly higher sEMG amplitude than plantarflexion in the posterior compartment in low, moderate, and high percentages of MVIC. However, no significant difference in sEMG amplitude was observed in the anterior compartment between eversion and plantarflexion. In addition, a posterior displacement of the COMx in eversion compared to plantarflexion in all MVIC percentages, with greater topographic distancing of the COMx at higher levels of activation. In conclusion, the peroneus longus muscle presented NMCs; the anterior compartment contributed to both eversion and plantarflexion movements, whereas the posterior compartment mainly contributed to the eversion movement of the ankle in low, moderate, and high percentages of MVIC.