EMG-based Pattern Recognition Research Articles

Research on the electromyography-based pattern recognition for inter-limb coordination in human crawling motion.

Aiming to provide a feasible crawling motion analysis method for clinical application, this study introduced electromyography (EMG)-based motion intention recognition technology into the pattern recognition of inter-limb coordination during human crawling for the first time. Eight inter-limb coordination modes (ILCMs) were defined. Ten adult participants were recruited, and each participant performed hands-knees crawling at low, medium, and fast speeds in self-selected ILCMs and the eight predefined ILCMs, respectively. EMG signals for pattern recognition were collected from 30 limbs and trunk muscles, and pressure signals for crawling cycle segmentation were collected from the left palm. The pattern recognition experiments were conducted in participant-specific, multi-participant, and participant-independent ways, respectively, adopting three different classifiers, including bidirectional long short-term memory (BiLSTM) network, support vector machine (SVM), and k-nearest neighbor (KNN). The experimental results show that EMG-based pattern recognition schemes could classify the eight ILCMs with high recognition rates, thereby confirming the feasibility of providing an EMG-based crawling motion analysis method for clinical doctors. Furthermore, based on the classification results of self-selected ILCMs at different speeds and the statistical results of stance duration, swing duration, and the duty factors of stance phase, the possible reasons why humans chose various ILCMs at different crawling speeds were discussed. The research results have potential application value for evaluating crawling function, understanding abnormal crawling control mechanisms, and designing rehabilitation robots.

HAND RECOGNITION SYSTEM BASED ON ELECTROMYOGRAPHY FOR REAL-TIME HAND PROSTHETIC CONTROL

Upper limb amputation is a significant limitation for achieving routine activities. Myoelectric signals detected by electrodes well-known as Electromyography (EMG) have been targeted to control upper limb prostheses of such lost limbs. Unfortunately, the acquisition, processing and use of such myoelectric signals are sophisticated. Furthermore, it necessarily requires complex computation to fulfil accuracy, robustness, and time-consumption execution for the real-time prosthesis application. Thus, machine learning schemes for pattern recognition are a potential approach to improve the traditional control for hand prostheses due to the movement of users and muscle contraction. This paper presents real-time hand posture recognition based on three hand postures using surface EMG (sEMG) signals. sEMG signals are acquired by the electrode channel and simultaneously collected while making a hand posture. Performance evaluation relies on classification accuracy and time consumption. The performance of six real-time recognition models is evaluated which combine two projection techniques and three classifiers. Results indicate that EMG-based pattern recognition (EMG-PR) control outperforms the traditional control for hand prostheses in real-time application. The highest classification accuracy is approximately 96%, whereas the lowest time consumption is 4 ms. In addition, the accuracy is dropped when the number of electrodes decreases nearly to 3%. These outcomes can apply to real-time hand prostheses to alleviate the limited prostheses available.

DLPR: Deep Learning-Based Enhanced Pattern Recognition Frame-Work for Improved Myoelectric Prosthesis Control

In EMG based pattern recognition (EMG-PR), deep learning-based techniques have become more prominent for their self-regulating capability to extract discriminant features from large data-sets. Moreover, the performance of traditional machine learning-based methods show limitation to categorize over a certain number of classes and degrades over a period of time. In this paper, an accurate, robust, and fast convolutional neural network-based framework for EMG pattern identification is presented. To assess the performance of the proposed system, five publicly available and benchmark data-sets of upper limb activities were used. This data-set contains 49 to 52 upper limb motions (NinaPro DB1, NinaPro DB2, and NinaPro DB3), Data with force variation, and data with arm position variation for intact and amputated subjects. The classification accuracies of 92.18% (53 classes), 91.56% (49 classes), 84.98% (49 classes of amputees), 95.67% (6 classes with force variation), and 99.12% (8 classes with arm position variation) have been observed during the testing. The performance of the proposed system is compared with the state of art techniques in the literature. The findings demonstrate that classification accuracy and time complexity have been improved significantly. For signal pre-processing and deep learning techniques, Keras which is a high-level API for TensorFlow is utilised to build profound learning models. The proposed method has been tested on the Intel Core i7 3.5GHz, 7th Gen CPU with 8GB DDR4 RAM.

An Exploration of the Optimal Feature-Classifier Combinations for Transradial Prosthesis Control.

Within state-of-the-art gesture-based upper-limb myoelectric prosthesis control, gesture recognition commonly relies on the classification of features extracted from electromyographic (EMG) data gathered from the amputee's residual forearm musculature. Despite best efforts in broadly maximizing gesture recognition accuracy, there does not yet exist a feature-classifier combination accepted as best-practice. In turn, this work hypothesizes that no single feature-classifier combination can consistently maximize accuracy across subjects, positing instead that control schemes should be personalized to the individual. To investigate this hypothesis, the study employed the 40-subject, 49-gesture Ninapro Database 2 (DB2) to compare the performance of 7 different historic, more recent, and state-of-the-art feature sets when classified by 5 machine learning algorithms commonly seen within EMG-based pattern recognition literature. The results demonstrate the ability of Linear Discriminant Analysis (LDA) to marginally exceed other more computationally intensive classifiers in terms of mean accuracy, while the feature set which maximized the highest proportion of individuals' accuracies was shown to vary with both classifier choice and gesture count.

Day-to-Day Stability of Wrist EMG for Wearable-Based Hand Gesture Recognition

Wrist electromyography (EMG) signals have been explored for incorporation into subtle wrist-worn wearable devices for decoding hand gestures. Previous studies have now shown that wrist EMG can even outperform the more commonly used forearm EMG, depending on the application. However, the performance and robustness of wrist EMG-based pattern recognition systems in the presence of confounding factors remain relatively unexplored. In this paper, we investigate the day-to-day stability of wrist EMG signals to ascertain their reliability across days. The test-retest reliability of concurrently collected wrist EMG and forearm EMG signals elicited during a variety of finger and wrist gestures was evaluated over a series of days. Several classification approaches, including a novel Maximum independence domain adaptation (MIDA), were investigated to explore and mitigate the effects of natural EMG variations across days. Results showed that wrist EMG signals were reliable and relatively resilient to the negative effects of EMG variations across days. Specifically, wrist EMG-based classifiers consistently outperformed forearm EMG-based classifiers with statistically significant differences ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p</i> < 0.05) and had average classification accuracies between 93.8% - 95.7% compared to 91.3% - 92.6% for the forearm EMG-based classifiers using a novel Inter-Day Feature Set (IDFS) and a novel adaptive-MIDA linear discriminant analysis (LDA) classification technique requiring minimal training. This study builds further evidence for the viability of commercial wrist-worn EMG wearables with minimal training burden for general consumers.

Feasibility of Two Different EMG-Based Pattern Recognition Control Paradigms to Control a Robot After Stroke - Case Study.

Stroke often results in chronic motor impairment of the upper-extremity yet neither traditional- nor robotics-based therapy has been able to affect this in a profound way. Supporting the weak affected shoulder against gravity improves reaching distance and minimizes abnormal co-contraction of the elbow, wrist, and fingers after stroke. However, it is necessary to assess the feasibility and efficacy of real-time controllers for this population as technology advances and a wearable shoulder device comes closer to reality. The aim of this study is to test two EMG-based controllers in this regard. A linear discriminant analysis based classifier was trained using extracted time domain and auto-regressive features from electromyographic data acquired during muscle effort required to move a load equivalent to 50 and 100% limb weight (abduction) and 150 and 200% limb weight (adduction). While rigidly connected to a custom lab-based robot, the participant was required to complete a series of lift and reach tasks under two different control paradigms: position-based control and force-based control. The participant successfully controlled the robot under both paradigms as indicated by first moving the robot arm into the proper vertical window and then reaching out as far as possible while remaining within the vertical window. This case study begins to assess the feasibility of using electromyographic data to classify the intended shoulder movement of a participant with stroke during a functional lift and reach type task. Next steps will assess how this type of support affects reaching function.

Causes of Performance Degradation in Non-invasive Electromyographic Pattern Recognition in Upper Limb Prostheses.

Surface Electromyography (EMG)-based pattern recognition methods have been investigated over the past years as a means of controlling upper limb prostheses. Despite the very good reported performance of myoelectric controlled prosthetic hands in lab conditions, real-time performance in everyday life conditions is not as robust and reliable, explaining the limited clinical use of pattern recognition control. The main reason behind the instability of myoelectric pattern recognition control is that EMG signals are non-stationary in real-life environments and present a lot of variability over time and across subjects, hence affecting the system's performance. This can be the result of one or many combined changes, such as muscle fatigue, electrode displacement, difference in arm posture, user adaptation on the device over time and inter-subject singularity. In this paper an extensive literature review is performed to present the causes of the drift of EMG signals, ways of detecting them and possible techniques to counteract for their effects in the application of upper limb prostheses. The suggested techniques are organized in a table that can be used to recognize possible problems in the clinical application of EMG-based pattern recognition methods for upper limb prosthesis applications and state-of-the-art methods to deal with such problems.

Zynq-based acceleration of robust high density myoelectric signal processing

Advances in electromyographic (EMG) sensor technology and machine learning algorithms have led to an increased research effort into high density EMG-based pattern recognition methods for prosthesis control. With the goal set on an autonomous multi-movement prosthesis capable of performing training and classification of an amputee’s EMG signals, the focus of this paper lies in the acceleration of the embedded signal processing chain. We present two Xilinx Zynq-based architectures for accelerating two inherently different high density EMG-based control algorithms. The first hardware accelerated design achieves speed-ups of up to 4.8 over the software-only solution, allowing for a processing delay lower than the sample period of 1 ms. The second system achieved a speed-up of 5.5 over the software-only version and operates at a still satisfactory low processing delay of up to 15 ms while providing a higher reliability and robustness against electrode shift and noisy channels.

A Robust Sparse Representation Based Pattern Recognition Approach for Myoelectric Control

Interferences in the form of white Gaussian noise (WGN) are inevitable during long-term electromyogram (EMG) recordings. Even with the aid of advanced signal denoising techniques, such an intermittent interference is hardly detected and attenuated in the practical use of EMG-driven control systems. Hence, a robust pattern recognition scheme that is invariant to noise contamination would potentially aid the realization of an efficient EMG-based pattern recognition (EMG-PR) control system. To this end, an EMG-PR scheme driven by sparse representation-based classification (SRC) algorithm and root mean square (rms) descriptor (RMS-SRC) is proposed in this paper. The accuracy and the robustness of the proposed scheme were investigated using the high-density surface EMG recordings from 12 traumatic braininjured patients and 5 post-stroke survivors. For benchmark comparison, another three different feature sets and four pattern recognition algorithms were considered. The optimal pattern recognition schemes with respect to each feature-classifier combination were first selected in the absence of WGN contamination. Then, six levels of WGN with a signal-to-noise ratio (SNR) ranging from 5 to 30 dB were added into the EMG recordings, respectively, to mimic the different WGN interferences. Our result showed that the proposed RMS-SRC scheme could achieve a similar accuracy with the benchmark schemes in the presence of limited noise contamination (0-15 dB), and significantly outperformed the other schemes when the SNR of WGN increased (20-30 dB). More notably, the RMS-SRC scheme significantly outperformed the other pattern recognition schemes when the WGN existed in either training set or testing set only. The findings proved the comparative advantage of the proposed RMS-SRC pattern recognition scheme over the other currently used schemes in the myoelectric control. Thus, the proposed scheme would potentially facilitate the development of EMG-driven rehabilitation robots for accurate and dexterous assistive training for patients with neurological disorders.

A strategy for labeling data for the neural adaptation of a powered lower limb prosthesis.

Pattern recognition algorithms that use EMG signals have been proposed to help control powered lower limb prostheses. These algorithms do not automatically compensate for disturbances in EMG signals, resulting in deterioration of algorithm accuracies. Supervised adaptive pattern recognition algorithms can solve this problem, but require correct labeling of new data. Information from embedded mechanical sensors can be compared to the characteristic gait profiles of the different modes to identify the mode of the user's most recent stride and provide a label for new data. The purpose of this study was to develop a gait pattern estimator (GPE) that could automatically make such a comparison. The GPE output was used to supervise an adaptive EMG-based pattern recognition algorithm. Our results indicate that using GPE-based adaptation helped prevent classification errors that would otherwise occur between experimental sessions. The GPE could accurately label new data with a low error rate of approx. 2%. The low error rate of the GPE was reflected in the accuracy of an adapted pattern recognition algorithm. The error rate of the adapted algorithm that was supervised by the GPE was not significantly different from one that used perfect supervision.

Electromyography-Based Locomotion Pattern Recognition and Personal Positioning Toward Improved Context-Awareness Applications

Personal positioning has been playing an important role in context awareness and navigation. Pedestrian dead reckoning (PDR) solution is a positioning technology used where the global positioning system (GPS) signal is not available or its signal is mightily attenuated or reflected by constructions nearby, such as inside the buildings or in GPS degraded areas such as urban city, basement. A traditional PDR solution employs a multisensor unit (integrating accelerometer, gyroscope, digital compass, barometer, etc.) to detect step occurrences, as well as to estimate the stride length. In our pilot research, we proposed a novel electromyography (EMG)-based method to fulfill that task and obtained satisfying PDR results. In this paper, a further attempt is made to investigate the feasibility of using EMG sensors in sensing muscle activities to detect the corresponding locomotion patterns, and as a result, a new approach, which recognizes different locomotion patterns using EMG signals and constructs stride length models according to the recognition results, is then proposed to improve the positioning accuracy and robustness of the EMG-based PDR solution by adapting the stride length model into different locomotion patterns. The experimental results demonstrate that EMG-based pattern recognition of four motions (walking, running, walking upstairs, walking downstairs) achieve an error rate of less than 2%. Combined with locomotion pattern recognition, the proposed EMG-based PDR solution yield a position deviation of less than 5 m within the whole distance of 404 m in a simulated indoor/outdoor field test. The proposed method is proven to be effective and practical in sensing context information, including both the user's activities and locations.

EMG-based pattern recognition approach in post stroke robot-aided rehabilitation: a feasibility study

BackgroundSeveral studies investigating the use of electromyographic (EMG) signals in robot-based stroke neuro-rehabilitation to enhance functional recovery. Here we explored whether a classical EMG-based patterns recognition approach could be employed to predict patients’ intentions while attempting to generate goal-directed movements in the horizontal plane.MethodsNine right-handed healthy subjects and seven right-handed stroke survivors performed reaching movements in the horizontal plane. EMG signals were recorded and used to identify the intended motion direction of the subjects. To this aim, a standard pattern recognition algorithm (i.e., Support Vector Machine, SVM) was used. Different tests were carried out to understand the role of the inter- and intra-subjects’ variability in affecting classifier accuracy. Abnormal muscular spatial patterns generating misclassification were evaluated by means of an assessment index calculated from the results achieved with the PCA, i.e., the so-called Coefficient of Expressiveness (CoE).ResultsProcessing the EMG signals of the healthy subjects, in most of the cases we were able to build a static functional map of the EMG activation patterns for point-to-point reaching movements on the horizontal plane. On the contrary, when processing the EMG signals of the pathological subjects a good classification was not possible. In particular, patients’ aimed movement direction was not predictable with sufficient accuracy either when using the general map extracted from data of normal subjects and when tuning the classifier on the EMG signals recorded from each patient.ConclusionsThe experimental findings herein reported show that the use of EMG patterns recognition approach might not be practical to decode movement intention in subjects with neurological injury such as stroke. Rather than estimate motion from EMGs, future scenarios should encourage the utilization of these signals to detect and interpret the normal and abnormal muscle patterns and provide feedback on their correct recruitment.