Electromyography Signal Processing Research Articles

Pattern Recognition in the Processing of Electromyographic Signals for Selected Expressions of Polish Sign Language.

Gesture recognition has become a significant part of human-machine interaction, particularly when verbal interaction is not feasible. The rapid development of biomedical sensing and machine learning algorithms, including electromyography (EMG) and convolutional neural networks (CNNs), has enabled the interpretation of sign languages, including the Polish Sign Language, based on EMG signals. The objective was to classify the game control gestures and Polish Sign Language gestures recorded specifically for this study using two different data acquisition systems: BIOPAC MP36 and MyoWare 2.0. We compared the classification performance of various machine learning algorithms, with a particular emphasis on CNNs on the dataset of EMG signals representing 24 gestures, recorded using both types of EMG sensors. The results (98.324% versus ≤7.8571% and 95.5307% versus ≤10.2697% of accuracy for CNNs and other classifiers in data recorded with BIOPAC MP36 and MyoWare, respectively) indicate that CNNs demonstrate superior accuracy. These results suggest the feasibility of using lower-cost sensors for effective gesture classification and the viability of integrating affordable EMG-based technologies into broader gesture recognition frameworks, providing a cost-effective solution for real-world applications. The dataset created during the study offers a basis for future studies on EMG-based recognition of Polish Sign Language.

Data-Driven Stroke Classification Utilizing Electromyographic Muscle Features and Machine Learning Techniques

Background: Predicting a stroke in advance or through early detection of subtle prodromal symptoms is crucial for determining the prognosis of the remaining life. Electromyography (EMG) has the advantage of easy and quick collection of biological data in clinical settings; however, its application in data processing and utilization is somewhat limited. Thus, this study aims to verify how simple signal processing and feature extraction utilize EMG in machine learning (ML)-based prediction models. Methods: EMG data were collected from the legs of 120 healthy individuals and 120 stroke patients during gait. Four statistical features were extracted from 16 EMG signals and trained on seven ML-based models. The accuracy of the validation and test datasets was also examined. Results: The model with the best performance was Random Forest. Among the 16 EMG signals, the average and maximum values of the muscle activities involved in knee extension (i.e., vastus medialis and rectus femoris) contributed significantly to the predictions. Conclusion: The results of this study confirmed that the simple processing and feature extraction of EMG signals effectively contributed to the accuracy of ML-based models. Routine use of EMG data collected in clinical environments is expected to provide benefits in terms of stroke prevention and rehabilitation.

Wearable sensor technology and medical robotics for fatigue assessment using electromyography signal processing

Wearable sensor technology and medical robotics for fatigue assessment using electromyography signal processing

Wearable device for personalized EMG feedback-based treatments

Curative effects of electromyography (EMG) feedback in treatment of various conditions and/or recovery after injuries have been earlier reported. However, wider application on EMG feedback is somehow limited due to the overall price of such systems and limited availability outside of the specialized treatment centers. Development of a personalized device for EMG feedback would be of great importance for home recovery after stroke or injuries, achieving better success in fitness and improving biofeedback-based treatments of conditions such as urinary incontinence. Despite extensive research on EMG signal collection, there is a lack of focus on in-situ EMG analysis that considers the intensity and duration of muscle activities. This gap presents the motivation for our research. In this paper, we present a methodology for the realization of wearable, rechargeable battery-powered, small-sized (90 mm × 60 mm) electronic device for recording two EMG channels (12-bits resolution, sampling frequency up to 1.6 kHz) with Bluetooth Low Energy connectivity to a smartphone. An average current consumption of 20.5 mA was experimentally determined, suggesting that multiday continuous functionality is possible. Advancing the state of the art, we propose a cross-correlation-based algorithm for in-situ dynamical computing and evaluation of muscle activation levels. This algorithm can determine if the muscle follows a predefined profile of contractions/relaxations (as needed for the treatment) and indicate if two muscles in a specific exercise were or were not engaged in proper time and with the proper intensity. The performed simulation analysis showed that the proposed approach exhibited shorter processing time compared to Morlet Wavelet Transform and Dynamic Time Warping. Finally, experimental work with five human volunteers demonstrated the reliability in EMG acquisition and signal processing. Therefore, the main contribution of this work is a cost-effective, small-sized, customizable EMG recording system with an efficient cross-correlation-based algorithm for in-situ EMG collection and processing.

Analysis of Electromyography (EMG) Signal Processing with Filtering Techniques

The paper presents the Analysis of Electromyography (EMG) Signal Processing with Filtering Techniques. The problem in this study is how to consider the filtering techniques for fundamental EMG signal processing with high-level accuracy. The research method for designing the simulation codes for observing the EMG signal modeling and digital filtering techniques with mathematical approaches from the signals and systems concepts. The results confirm that the outcomes of this study met the performance target for noise removal techniques of EMG signals in real-world applications.

Prosthetic hand for phocomelia patients with enhanced pain management using EMG signal processing and biofeedback control system

Denoising is an essential step in data mining. It makes an effort to remove noise from the picture without sacrificing any of the important elements. The primary method utilized in this project to assess the effectiveness of phocomelia images is Kmeans clustering. The PSNR and MSE values of the denoised image are computed. In the end, the best technique for denoising medical images is chosen based on the PSNR values from the collection of patient data necessary for the fitting of upper limb devices. Using image processing, the bone health of phocomelia patients are evaluated and specific information extracted from MRI scans. Electromyography (EMG) sensors pick up electrical impulses, convert them into hand movements that the user wants, and flex the muscles in the residual limb directly below the elbow. The same muscles that enable hand function in humans are also felt by the system using microcontrollers. Through the integration of sophisticated sensor technologies with responsive prosthetic design idea, this newly developed technique enhances feedback from the proprioceptive system for patients with phocomelia, hence promoting natural movement and improving functional outcomes. As a result, it provides phocomelia patients a special method for acquiring superior result from biofeedback system.

An Introduction to Electromyography Signal Processing and Machine Learning for Pattern Recognition: A Brief Overview

Electromyography (EMG) is about studying electrical signals from muscles and can provide a wealth of information on the function, contraction, and activity of your muscles. In the field of EMG pattern recognition, these signals are used to identify and categorize patterns linked to muscle activity. Various machine learning (ML) methods are used for this purpose. Successful detection of these patterns depends on using effective signal-processing techniques. It is crucial to reduce noise in EMG for accurate and meaningful information about muscle activity, improving signal quality for precise assessments. ML tools such as SVMs, neural networks, KNNs, and decision trees play a crucial role in sorting out complex EMG signals for different pattern recognition tasks. Clustering algorithms also help analyze and interpret muscle activity. EMG and ML find diverse uses in rehabilitation, prosthetics, and human-computer interfaces, though real-time applications come with challenges. They bring significant changes to prosthetic control, human-computer interfaces, and rehabilitation, playing a vital role in pattern recognition. They make prosthetic control more intuitive by understanding user intent from muscle signals, enhance human-computer interaction with responsive interfaces, and support personalized rehabilitation for those with motor impairments. The combination of EMG and ML opens doors for further research into understanding muscle behavior, improving feature extraction, and advancing classification algorithms.

A 3-D-Printed Portable EMG Wristband for the Quantitative Detection of Finger Motion

It is highly required to develop an affordable surface electromyography (EMG) wristband for the improved control of prostheses that is suitable to mimic human hand functions. By using the 3D-printing technology, the wristband is customizable and feasible for various arm sizes or shapes. In this paper, we have developed the 3D printed wristband for the control application. It consists of five pairs of custom-made serpentine dry electrodes, one EMG sensor, and one signal processing printed circuit board (PCB) to detect the user’s finger movements simultaneously. The wristband demonstrated its stability in EMG signal processing with a single-sensor system, which differed from other multi-sensor system devices. We verified the stability of the serpentine electrodes as bendable and stretchable. The serpentine electrodes have improved signal detection by providing conformal contact to the skin surface compared with the existing rigid dry electrodes. The flexibility allowed the electrodes to be placed in any shape of the skin surface. In a series of tests performed by the volunteer, we showed that the collected EMG signals reflected the muscle movements through signal processing. Under the muscle contraction of 75 lb, the wristband showed a signal sensitivity of 0.556 mV/lb with a 27 dB signal-to-noise ratio (SNR) for its valid EMG sensing capability. In addition, we also demonstrated the relationship between signal intensities and muscle forces at the different levels quantitatively. This work shows promising potential toward the advanced control system of prosthetic fields and its corresponding market.

A Real-Time EMG-Based Fixed-Bandwidth Frequency-Domain Embedded System for Robotic Hand.

The signals from electromyography (EMG) have been used for volitional control of robotic assistive devices with the challenges of performance improvement. Currently, the most common method of EMG signal processing for robot control is RMS (root mean square)-based algorithm, but system performance accuracy can be affected by noise or artifacts. This study hypothesized that the frequency bandwidths of noise and artifacts are beyond the main EMG signal frequency bandwidth, hence the fixed-bandwidth frequency-domain signal processing methods can filter off the noise and artifacts only by processing the main frequency bandwidth of EMG signals for robot control. The purpose of this study was to develop a cost-effective embedded system and short-time Fourier transform (STFT) method for an EMG-controlled robotic hand. Healthy volunteers were recruited in this study to identify the optimal myoelectric signal frequency bandwidth of muscle contractions. The STFT embedded system was developed using the STM32 microcontroller unit (MCU). The performance of the STFT embedded system was compared with RMS embedded system. The results showed that the optimal myoelectric signal frequency band responding to muscle contractions was between 60 and 80 Hz. The STFT embedded system was more stable than the RMS embedded system in detecting muscle contraction. Onsite calibration was required for RMS embedded system. The average accuracy of the STFT embedded system is 91.55%. This study presents a novel approach for developing a cost-effective and less complex embedded myoelectric signal processing system for robot control.

A Technique for Surface Electromyography Signal Processing Used for Robotic Device Control

A Technique for Surface Electromyography Signal Processing Used for Robotic Device Control

EMG-driven fatigue-based self-adapting admittance control of a hand rehabilitation robot

Upper-limb rehabilitation therapy sessions for post-stroke people generally contain rhythmic hand movements in a tiresome manner to rebuild the injured neural circuits. Fatigue formation causes breaks in the training and limits the therapy duration. Therefore, it is essential to establish a correlation between the patient’s muscle condition and the rehabilitation exercises to improve the physiotherapy sessions. A self-adapting control method based on online fatigue detection in rhythmic arm movements is presented. Experimental tests were performed on twenty healthy subjects to validate the method’s feasibility. Electromyography (EMG) and force signals considered the interfaces between users and the robot. In the first stage of the experiment, utilizing the frequency features from EMG signals, a neural network for fatigue detection trained; however, in the end, it substituted with a simple function as a refinement in the time-consuming aspect for the online employment. The initiation of the fatigue process is followed by reducing the admittance controller damping term based on the EMG signal processing. Trajectory tracking with the robot employs the self-adapting admittance controller (SAAC) method and the non-adapting admittance controller (NAAC). Movement accuracy and smoothness were measured and showed a better performance of the SAAC method related to the NAAC. Simulations with two different stiffness levels were performed on an upper-limb OpenSim model to study a stroke-injured arm and evaluate the proposed method’s proficiency. The metabolic cost indicated the movement’s superiority in a fatigue situation for reduced environment stiffness.

A Review on EMG Signal Classification and Applications

Electromyography (EMG) signals are muscles signals that enable the identification of human movements without the need of complex human kinematics calculations. Researchers prefer EMG signals as input signals to control prosthetic arms and exoskeleton robots. However, the proper algorithm to classify human movements from raw EMG signals has been an interesting and challenging topic to researchers. Various studies have been carried out to produce EMG-based human movement classification that gives high accuracy and high reliability. In this paper, the methods used in EMG signal acquisition and processing are reviewed. The different types of feature extraction techniques preferred by researchers are also discussed, including some combination and comparison of feature extraction techniques. This paper also reviews the different types of classifiers favored by researchers to recognize human movements based on EMG signals. The current applications of EMG signals are also reviewed.

Discussion of the Influence of Multiscale PCA Denoising Methods with Three Different Features.

Bioinformation is information generated from biological movement. By using a variety of modern technologies, we can use this information to form a meaningful model for researchers to study. An electromyographic (EMG) signal is one type of bioinformation that is used in many areas to help people study human muscle movement. This information can help in both clinical areas and industrial areas. EMG is a very complicated signal, so processing it is vital. The processing of EMG signals is divided into collection, denoising, decomposition, feature extraction and classification steps. In this article, the wavelet denoising step and several decomposition processes are discussed to show the usage of this technique in the final classification step. At the end of the study, we find that after the wavelet denoising step, the classification accuracy, which uses the K-nearest neighbor of the independent component analysis features, improves, but the accuracy of the wavelet coefficient features and autoregression coefficient features decreases.

Electromechanical Hand Prototype for the Simulation of the Opening and Closing Movement

This work aims to describe the construction of a low-cost functional prototype of an electromechanical hand with opening and closing capacity. The device, built on an ABS filament extruder, is actuated by a MG995 high torque servo motor, controlled by an Arduino-UNO R3 board, from the processing of electromyographic signals carried out by: MyoWare Muscle Sensor. The realization and functional evaluation of the device is developed in six phases: The first is to select the thermo-deformable materials, the second is to select the electronic components; the third phase is the design of the electromechanical hand prosthesis, the fourth phase was the construction of the mechanical parts; the fifth phase is the development of the control program on the electronic card; and the sixth phase of the project consists of validating the developed prototype by observing its opening and closing through the oscilloscope. Due to the low cost of development and ease of reproducibility (because it is built under the open-source philosophy), this work is expected to make a contribution to the disabled community, especially the one with limited resources.

Processing of EMG Signals with High Impact of Power Line and Cardiac Interferences

This work deals with electromyography (EMG) signal processing for the diagnosis and therapy of different muscles. Because the correct muscle activity measurement of strongly noised EMG signals is the major hurdle in medical applications, a raw measured EMG signal should be cleaned of different factors like power network interference and ECG heartbeat. Unfortunately, there are no completed studies showing full multistage signal processing of EMG recordings. In this article, the authors propose an original algorithm to perform muscle activity measurements based on raw measurements. The effectiveness of the proposed algorithm for EMG signal measurement was validated by a portable EMG system developed as a part of the EU research project and EMG raw measurement sets. Examples of removing the parasitic interferences are presented for each stage of signal processing. Finally, it is shown that the proposed processing of EMG signals enables cleaning of the EMG signal with minimal loss of the diagnostic content.

Review on electromyography signal acquisition and processing.

Electromyography (EMG) is a technique for recording biomedical electrical signals obtained from the neuromuscular activities. These signals are used to monitor medical abnormalities and activation levels, and also to analyze the biomechanics of any animal movements. In this article, we provide a short review of EMG signal acquisition and processing techniques. The average efficiency of capture of EMG signals with current technologies is around 70%. Once the signal is captured, signal processing algorithms then determine the recognition accuracy, with which signals are decoded for their corresponding purpose (e.g., moving robotic arm, speech recognition, gait analysis). The recognition accuracy can go as high as 99.8%. The accuracy with which the EMG signal is decoded has already crossed 99%, and with improvements in deep learning technology, there is a large scope for improvement in the design hardware that can efficiently capture EMG signals.

The difference of surface electromyography data processing method based on simulated manal-lifting-task

Objective: To study the differences of different signal processing method of surface electromyography (sEMG) in judging muscle fatigue. Methods: From July to October 2019, based on the model of simulated manual lifting operation, the original sEMG signals from 13 volunteers of brachial radial muscle, brachial two-headed muscle, triangle muscle, left vertical spine muscle, right vertical spine muscle and lateral femoral muscle were collected in the operation activities. Three different electromyography signal processing methods (all signal from motion beginning to the end, peak signal and ehe specified motion signal) were used to analyze the original data in time domain (RMS) and frequency domain (MDF) , the data difference between different electromyography signal processing methods was analyzed by using Wilcoxon rank and sum test and nonlinear curve fitting method. Results: The age of the subjects of the simulated lifting operation was (24.31±2.02) years old, height (173.78±4.84) cm, weight (66.28±5.58) kg, body mass index (BMI) 21.94±1.58. The thickness of triceps skinfold was (14.08±4.86) mm, and the thickness of the skin fold under the scapula was (15.54±3.59) mm. After processing the original signal data by using different sEMG signal interception methods, the normality test, Levene's test, and the Wilcoxon test showed that, except for the MDF index of the brachial two-headed muscle, the differences in the RMS and MDF signals of the other muscles were statistically significant (P<0.016) . The all signal processing method dealed with data distribution dispersion better than other methods, and the rate of change of RMS signal slope was higher than other methods. Non-linear regression results showed that all signal processing method had low volatility in processing data, and the regression equation had a high degree of fit. Conclusion: Different electromyography signal processing methods have differences. The all signal processing method which intercepts from starting point to the end point of action cycle has the least data volatility, and electromyography time domain and frequency domain index with the highest sensitivity of time, which is suitable for the application of surface electromyography to judge muscle fatigue in dynamic and complex operations.

Adaptive motion intent understanding–based control of human–exoskeleton system

Although human–exoskeleton systems have tremendous potentials for rehabilitation training, transparent human–exoskeleton interaction has not been achieved for late-phase rehabilitation. To cope with this challenge, an adaptive pilot intent prediction–based control algorithm is proposed to achieve compliant motion coordination. A novel surface electromyography signal processing method providing definite physical meaning and high computational efficiency is derived. The state space model of human musculoskeletal system is developed with the input of the processed surface electromyography. Adaptive parameter estimation is employed to cope with the emerging dynamics. It integrates the advantages of surface electromyography signals which imply muscle contraction in advance and force signals with high stability. Based on the coupled human–exoskeleton system dynamics, transparent control of a human–exoskeleton system is realized. Experimental results show that the computational efficiency of the proposed surface electromyography processing method is 129% higher than the energy kernel method, and the maximum interactive force is reduced about 15 N compared with impedance control.

The Method of the Main Tone Detection in the Structure of Electromyographic Signals for the Task of Broken Human Communicative Function Compensation

In the work the method of electromyographic signals processing for the task of broken human communicative function compensation is developed. The method allows to detect the signs of main tone in the structure of electromyographic signals, that were recorded from the surface of patients neck near the vocal folds. Using this signs it is possible to identify the mentally spoken vowel and vocalised consonant phonemes and to identify the speech of patients with broken or lost communicative function. The developed method includes two stages, namely: preparatory and basic. The purpose of the preparatory stage is to obtain data on the individual features of the patient's speach, in particular the approximate value of the main tone frequency and the frequency interval of the main tone frequency when patient is trying to utter test sequences of sounds at certain points in time. These data are necessary to enable the use of the basic stage of the method, which involves the processing of electromyographic signals recorded in an arbitrary attempt to pronounce arbitrary sounds, words or phrases by the patient. It is proposed to processing the electromyographic signals by methods of spectral-correlation analysis using the sliding window method if presenting of such biosignals in the form of a piecewise stationary random process to detect the time intervals of the presence of main tone signs. Within each sliding window the estimates of the power spectral density distribution are calculated and averaged over the frequency and power within the predetermined interval of existence of the main tone frequency. The obtained averaged estimates make it possible to set the time intervals of the main tone presence and, accordingly, the subsequent identification of vowel and consonant vocalised phonemes. The experimentally registered EMG signal was processed by the developed method with different values of the sliding window width.

Review on Electromyography Signal Acquisition, Processing and Its Applications

Electromyography (EMG) signal is the type of biomedical signal, which is obtained from the neuromuscular activities. Typically, an electromyogram instrument is used to capture the EMG signals. These signals are used to monitor medical abnormalities, activation level, and also to analyze the biomechanics of any animal movements. In this current work, we provide a short review of EMG signal acquisition and processing techniques. We found that the average efficiency to capture EMG signals with the current technologies is around 70 %. Once the signal is captured, the signal processing algorithms applied decides the recognition accuracy, with which signals are decoded for their corresponding purpose (e.g. moving robotic arm, speech recognition, gait analysis, etc). The recognition accuracy can go as high as 99.8 %. The accuracy with which the EMG signal is decoded has already crossed 99 %, and with the upcoming deep learning technology, there is a scope of improvement to design hardware, that can efficiently capture EMG signals.