Abstract:PurposeLoss of movement in the upper limb is a limitation that affects the lives of many individuals. This problem motivates the creation of several solutions, from intelligent biomechanical prostheses to new physical therapy approaches. However, one of the main limitations of these applications is in the detection and recognition of signal patterns for movement classification, considering both accuracy and computational complexity. The main hypothesis of this work is that extreme learning machines and echo state machines can combine acceptable accuracy with the low memory usage and processing typical of these machine learning models, in order to promote greater autonomy to users. In this work, we proposed a simple method for electromyography (EMG) classification, disregarding feature selection and post-processing algorithms.MethodIn this work, we propose the use of extreme learning machines and echo state machines as a low computational cost alternative for motion recognition from myoelectric signals. Other classifier models, such as the multilayer perceptron and linear discriminant analysis, were investigated. To validate the proposal, we used two datasets that represent two scenarios: an ideal environment, with no data on amputees; and a large dataset with myoelectric signals from patients with amputation. The results were evaluated as a function of accuracy and kappa index.ResultsUsing an extreme learning machine, the results for the dataset of healthy individuals reached 99.10% accuracy, with a processing time per instance of 0.09 ms. Considering the amputee dataset, the best results were obtained with the use of extreme learning machine augmentation, with an accuracy of 75.60% and a kappa of 0.7507.ConclusionWe showed that Extreme Learning Machines without pre- and post-data processing is feasible for real-time application, presenting good performance and increased velocity when compared to usual algorithms. Also, data augmentation appears to be a good approach for robustly classifying unbalanced EMG datasets.

Comparison of machine learning algorithms and feature extraction techniques for the automatic detection of surface EMG activation timing

Comparison of Bagging and Boosting Ensemble Machine Learning Methods for Automated EMG Signal Classification

Machine Learning for Detection of Muscular Activity from Surface EMG Signals

Automatic detection of surface EMG activation timing using a wavelet transform based method

Denoising Surface Electromyography Signals Based on Complementary Ensemble Empirical-Mode Decomposition and Multi-Scale Entropy

A novel feature extraction method for machine learning based on surface electromyography from healthy brain

Prediction of Lower Limb Action Intention Based on Surface EMG Signal

Distinguishing normal, neuropathic and myopathic EMG with an automated machine learning approach

Surface electromyography classification using extreme learning machines and echo state networks

MACHINE LEARNING TECHNIQUES APPLIED IN SURFACE EMG DETECTION- A SYSTEMATIC REVIEW

Hand Movement Detection from Surface Electromyography Signals by Machine Learning Techniques

Overview of processing techniques for surface electromyography signals

Implementation of an Intelligent EMG Signal Classifier Using Open-Source Hardware

Innovative use of optical sensors for real-time imaging and diagnosis using enhanced verge denoising and LSTM for sports medicine

An Investigation into Mechanomyography for Signal Extraction and Classification of Human Lower Limb Activity

Variational Mode Decomposition and Empirical Wavelet Transform-Based Feature Extraction and Ensemble Classifier for Lower Limb Movement Prediction With Surface Electromyography Signal

Surface EMG Feature Disentanglement for Robust Pattern Recognition.

MSEMG: Surface Electromyography Denoising with a Mamba-based Efficient Network

Extraction and analysis of multiple time window features associated with muscle fatigue conditions using sEMG signals

A machine learning approach to identify hand actions from single-channel sEMG signals

Detecting Psychological Interventions Using Bilateral Electromyographic Wearable Sensors