Surface electromyogram signal modelling

Recognition of Hand Gestures Based on EMG Signals with Deep and Double-Deep Q-Networks

In recent years, hand gesture recognition (HGR) technologies that use electromyography (EMG) signals have been of considerable interest in developing human-machine interfaces. Most state-of-the-art HGR approaches are based mainly on supervised machine learning (ML). However, the use of reinforcement learning (RL) techniques to classify EMGs is still a new and open research topic. Methods based on RL have some advantages such as promising classification performance and online learning from the user's experience. In this work, we propose a user-specific HGR system based on an RL-based agent that learns to characterize EMG signals from five different hand gestures using Deep Q-network (DQN) and Double-Deep Q-Network (Double-DQN) algorithms. Both methods use a feed-forward artificial neural network (ANN) for the representation of the agent policy. We also performed additional tests by adding a long-short-term memory (LSTM) layer to the ANN to analyze and compare its performance. We performed experiments using training, validation, and test sets from our public dataset, EMG-EPN-612. The final accuracy results demonstrate that the best model was DQN without LSTM, obtaining classification and recognition accuracies of up to 90.37%±10.7% and 82.52%±10.9%, respectively. The results obtained in this work demonstrate that RL methods such as DQN and Double-DQN can obtain promising results for classification and recognition problems based on EMG signals.

Toward a generalizable deep CNN for neural drive estimation across muscles and participants

Objective.High-density electromyography (HD-EMG) decomposition algorithms are used to identify individual motor unit (MU) spike trains, which collectively constitute the neural code of movements, to predict motor intent. This approach has advanced from offline to online decomposition, from isometric to dynamic contractions, leading to a wide range of neural-machine interface applications. However, current online methods need offline retraining when applied to the same muscle on a different day or to a different person, which limits their applications in a real-time neural-machine interface. We proposed a deep convolutional neural network (CNN) framework for neural drive estimation, which takes in frames of HD-EMG signals as input, extracts general spatiotemporal properties of MU action potentials, and outputs the number of spikes in each frame. The deep CNN can generalize its application without retraining to HD-EMG data recorded in separate sessions, muscles, or participants.Approach.We recorded HD-EMG signals from the vastus medialis and vastus lateralis muscles from five participants while they performed isometric contractions during two sessions separated by ∼20 months. We identified MU spike trains from HD-EMG signals using a convolutive blind source separation (BSS) method, and then used the cumulative spike train (CST) of these MUs and the HD-EMG signals to train and validate the deep CNN.Main results.On average, the correlation coefficients between CST from the BSS and that from deep CNN were0.983±0.006for leave-one-out across-sessions-and-muscles validation and0.989±0.002for leave-one-out across-participants validation. When trained with more than four datasets, the performance of deep CNN saturated at0.984±0.001for cross validations across muscles, sessions, and participants.Significance.We can conclude that the deep CNN is generalizable across the aforementioned conditions without retraining. We could potentially generate a robust deep CNN to estimate neural drive to muscles for neural-machine interfaces.

A fused biometrics information graph convolutional neural network for effective classification of patellofemoral pain syndrome

Patellofemoral pain syndrome (PFPS) is a common, yet misunderstood, knee pathology. Early accurate diagnosis can help avoid the deterioration of the disease. However, the existing intelligent auxiliary diagnosis methods of PFPS mainly focused on the biosignal of individuals but neglected the common biometrics of patients. In this paper, we propose a PFPS classification method based on the fused biometrics information Graph Convolution Neural Networks (FBI-GCN) which focuses on both the biosignal information of individuals and the common characteristics of patients. The method first constructs a graph which uses each subject as a node and fuses the biometrics information (demographics and gait biosignal) of different subjects as edges. Then, the graph and node information [biosignal information, including the joint kinematics and surface electromyography (sEMG)] are used as the inputs to the GCN for diagnosis and classification of PFPS. The method is tested on a public dataset which contain walking and running data from 26 PFPS patients and 15 pain-free controls. The results suggest that our method can classify PFPS and pain-free with higher accuracy (mean accuracy = 0.8531 ± 0.047) than other methods with the biosignal information of individuals as input (mean accuracy = 0.813 ± 0.048). After optimal selection of input variables, the highest classification accuracy (mean accuracy = 0.9245 ± 0.034) can be obtained, and a high accuracy can still be obtained with a 40% reduction in test variables (mean accuracy = 0.8802 ± 0.035). Accordingly, the method effectively reflects the association between subjects, provides a simple and effective aid for physicians to diagnose PFPS, and gives new ideas for studying and validating risk factors related to PFPS.

Automated semi-real-time detection of muscle activity with ultrasound imaging

Ultrasound imaging (USI) biofeedback is a useful therapeutic tool; however, it relies on qualitative assessment by a trained therapist, while existing automatic analysis techniques are computationally demanding. This study aims to present a computationally inexpensive algorithm based on the difference in pixel intensity between USI frames. During an offline experiment, where data was analyzed after the study, participants performed isometric contractions of the gastrocnemius medialis (GM) muscle, as executed (30% of maximum contraction) or attempted (low force contraction up to a point when the participant is aware of exerting force or contracting the muscle) movements, while USI, EMG, and force data were recorded. The algorithm achieved 99% agreement with EMG and force measurements for executed movements and 93% for attempted movements, with USI detecting 1.9% more contractions than the other methods. In the online study, participants performed GM muscle contractions at 10% and 30% of maximum contraction, while the algorithm provided visual feedback proportional to the muscle activity (based on USI recordings during the maximum contraction) in less than 3 s following each contraction. We show that the participants reached the target consistently, learning to perform precise contractions. The algorithm is reliable and computationally very efficient, allowing real-time applications on standard computing hardware. It is a suitable method for automated detection, quantification of muscle contraction, and to provide biofeedback which can be used for training of targeted muscles, making it suitable for rehabilitation.

A New Muscle Activation Dynamics Model, That Simulates the Calcium Kinetics and Incorporates the Role of Store-Operated Calcium Entry Channels, to Enhance the EMG-Driven Hill-type Models

Hill-type models are frequently used in biomechanical simulations. They are attractive for their low computational cost and close relation to commonly measured musculotendon parameters. Still, more attention is needed to improve the activation dynamics of the model specifically because of the nonlinearity observed in the EMG-Force relation. Moreover, one of the important and practical questions regarding the assessment of the model's performance is how adequately can the model simulate any fundamental type of human movement without modifying model parameters for different tasks? This paper tries to answer this question by proposing a simple physiologically based activation dynamics model. The model describes the ?kinetics of the calcium dynamics while activating and deactivating the muscle contraction process. Hence, it allowed simulating the recently discovered role of store-operated calcium entry (SOCE) channels as immediate counter-flux to calcium loss across the tubular system during excitation-contraction coupling. By comparing the ability to fit experimental data without readjusting the parameters, the proposed model has proven to have more steady performance than phenomenologically based models through different submaximal isometric contraction levels. This model indicates that more physiological insights is key for improving Hill-type model performance.

Gesture Recognition Using Surface Electromyography and Deep Learning for Prostheses Hand: State-of-the-Art, Challenges, and Future

Amputation of the upper limb brings heavy burden to amputees, reduces their quality of life, and limits their performance in activities of daily life. The realization of natural control for prosthetic hands is crucial to improving the quality of life of amputees. Surface electromyography (sEMG) signal is one of the most widely used biological signals for the prediction of upper limb motor intention, which is an essential element of the control systems of prosthetic hands. The conversion of sEMG signals into effective control signals often requires a lot of computational power and complex process. Existing commercial prosthetic hands can only provide natural control for very few active degrees of freedom. Deep learning (DL) has performed surprisingly well in the development of intelligent systems in recent years. The significant improvement of hardware equipment and the continuous emergence of large data sets of sEMG have also boosted the DL research in sEMG signal processing. DL can effectively improve the accuracy of sEMG pattern recognition and reduce the influence of interference factors. This paper analyzes the applicability and efficiency of DL in sEMG-based gesture recognition and reviews the key techniques of DL-based sEMG pattern recognition for the prosthetic hand, including signal acquisition, signal preprocessing, feature extraction, classification of patterns, post-processing, and performance evaluation. Finally, the current challenges and future prospects in clinical application of these techniques are outlined and discussed.

Real-Time Hand Gesture Recognition Using Surface Electromyography and Machine Learning: A Systematic Literature Review

Today, daily life is composed of many computing systems, therefore interacting with them in a natural way makes the communication process more comfortable. Human-Computer Interaction (HCI) has been developed to overcome the communication barriers between humans and computers. One form of HCI is Hand Gesture Recognition (HGR), which predicts the class and the instant of execution of a given movement of the hand. One possible input for these models is surface electromyography (EMG), which records the electrical activity of skeletal muscles. EMG signals contain information about the intention of movement generated by the human brain. This systematic literature review analyses the state-of-the-art of real-time hand gesture recognition models using EMG data and machine learning. We selected and assessed 65 primary studies following the Kitchenham methodology. Based on a common structure of machine learning-based systems, we analyzed the structure of the proposed models and standardized concepts in regard to the types of models, data acquisition, segmentation, preprocessing, feature extraction, classification, postprocessing, real-time processing, types of gestures, and evaluation metrics. Finally, we also identified trends and gaps that could open new directions of work for future research in the area of gesture recognition using EMG.

An Online Spectral Information-Enhanced Approach for Artifact Detection and Fault Attentuation in Myoelectric Control

In myocontrol of neuroprosthetic devices, multichannel electromyography (EMG) can be used to decode the intended motor command, based on distributed activation patterns of stump muscles. In this regard, the high density EMG (HD-EMG) approach allows for enhancement of the spatiotemporal resolution for motor intention detection. Despite the advantages of relying on several EMG channels, the challenge of high-density electrode systems is the dynamically changing electrode-skin contact impedance, which can affect a considerable number of electrodes over the time of data acquisition. This can result in obtaining unreliable, low-quality EMG recording with a distributed artifact pattern over the grid of EMG sensors. To address this issue, we propose a novel online approach for adaptive information extraction and enhancement for automatic artifact detection and attenuation in HD-EMG-based myocontrol of prosthetic devices. The method is based on an adaptive weighting scheme that modifies the contribution of each HD-EMG channel considering the spectral information content relative to artifacts. The technique (named IE-HD-EMG) was tested as an online pre-conditioning step for a challenging multiclass classification problem of 4-finger activation, using linear discriminant analysis. It is shown that for this application, the proposed IE-HD-EMG technique led to a superior performance in finger activation recognition (79.25% accuracy, 89% sensitivity, 89.15% specificity) in comparison to the conventional HD-EMG recording under the same condition without the proposed approach (56.25% accuracy, 61.3% sensitivity, 67% specificity). Therefore, the proposed technique can have a significant potential to expand the clinical viability of HD-EMG systems.

Head magnetomyography (hMMG): A novel approach to monitor face and whole head muscular activity

Muscular activity recording is of high basic science and clinical relevance and is typically achieved using electromyography (EMG). While providing detailed information about the state of a specific muscle, this technique has limitations such as the need for a priori assumptions about electrode placement and difficulty with recording muscular activity patterns from extended body areas at once. For head and face muscle activity, the present work aimed to overcome these restrictions by exploiting magnetoencephalography (MEG) as a whole head myographic recorder (head magnetomyography, hMMG). This is in contrast to common MEG studies, which treat muscular activity as artifact in electromagnetic brain activity. In a first proof‐of‐concept step, participants imitated emotional facial expressions performed by a model. Exploiting source projection algorithms, we were able to reconstruct muscular activity, showing spatial activation patterns in accord with the hypothesized muscular contractions. Going one step further, participants passively observed affective pictures with negative, neutral, or positive valence. Applying multivariate pattern analysis to the reconstructed hMMG signal, we were able to decode above chance the valence category of the presented pictures. Underlining the potential of hMMG, a searchlight analysis revealed that generally neglected neck muscles exhibit information on stimulus valence. Results confirm the utility of hMMG as a whole head electromyographic recorder to quantify muscular activation patterns including muscular regions that are typically not recorded with EMG. This key advantage beyond conventional EMG has substantial scientific and clinical potential.

We present an innovative method called head magnetomyography (hMMG), which exploits magnetoencephalography (MEG) as a whole head electromyographic (EMG) recorder. Differently from the typical EMG recording, which needs an a priori selection of the placement of the electrodes, hMMG is able to detect muscular activity from many regions of the face and head simultaneously, including typically overlooked muscles. Our data show that hMMG can readily serve researchers in the emotion field and hold further scientific as well as clinical promise.

A Novel Validation Approach for High-Density Surface EMG Decomposition in Motor Neuron Disease

This paper presents a novel two-source approach for validating the performance of high-density surface electromyogram (EMG) decomposition. The approach was developed taking advantage of surface EMG characteristics of amyotrophic lateral sclerosis (ALS). High-density surface EMG data from ALS patients can be divided to the sparse data set and the interference data set, with the former decomposed by expert visual inspection while the latter independently decomposed by the surface EMG decomposition algorithm. The agreement of the decomposition yields from the two data sets can be quantified for evaluating the surface EMG decomposition performance. The novel validation approach was performed for a recently developed method called automatic progressive FastICA peel-off (APFP) for high-density surface EMG decomposition. The APFP framework was used to automatically decompose high-density surface EMG signals recorded from the first dorsal interosseous muscle of ALS subjects. The common motor units independently decomposed from the interference data set and the sparse data set demonstrated an average matching rate of 99.18% ± 1.11%. The characteristics of the ALS surface EMG also facilitate a step by step illustration of the APFP framework for high-density surface EMG decomposition. The novel approach presented in this paper can supplement conventional two-source validation for accuracy assessment of decomposed motor units from experimental signals, which is essential for development of surface EMG decomposition methods.

Non-invasive Decoding of the Motoneurons: A Guided Source Separation Method Based on Convolution Kernel Compensation With Clustered Initial Points

Despite the progress in understanding of neural codes, the studies of the cortico-muscular coupling still largely rely on interferential electromyographic (EMG) signal or its rectification for the assessment of motor neuron pool behavior. This assessment is non-trivial and should be used with precaution. Direct analysis of neural codes by decomposing the EMG, also known as neural decoding, is an alternative to EMG amplitude estimation. In this study, we propose a fully-deterministic hybrid surface EMG (sEMG) decomposition approach that combines the advantages of both template-based and Blind Source Separation (BSS) decomposition approaches, a.k.a. guided source separation (GSS), to identify motor unit (MU) firing patterns. We use the single-pass density-based clustering algorithm to identify possible cluster representatives in different sEMG channels. These cluster representatives are then used as initial points of modified gradient Convolution Kernel Compensation (gCKC) algorithm. Afterwards, we use the Kalman filter to reduce the noise impact and increase convergence rate of MU filter identification by gCKC. Moreover, we designed an adaptive soft-thresholding method to identify MU firing times out of estimated MU spike trains. We tested the proposed algorithm on a set of synthetic sEMG signals with known MU firing patterns. A grid of 9 × 10 monopolar surface electrodes with 5-mm inter-electrode distances in both directions was simulated. Muscle excitation was set to 10, 30, and 50%. Colored Gaussian zero-mean noise with the signal-to-noise ratio (SNR) of 10, 20, and 30 dB, respectively, was added to 16 s long sEMG signals that were sampled at 4,096 Hz. Overall, 45 simulated signals were analyzed. Our decomposition approach was compared with gCKC algorithm. Overall, in our algorithm, the average numbers of identified MUs and Rate-of-Agreement (RoA) were 16.41 ± 4.18 MUs and 84.00 ± 0.06%, respectively, whereas the gCKC identified 12.10 ± 2.32 MUs with the average RoA of 90.78 ± 0.08%. Therefore, the proposed GSS method identified more MUs than the gCKC, with comparable performance. Its performance was dependent on the signal quality but not the signal complexity at different force levels. The proposed algorithm is a promising new offline tool in clinical neurophysiology.

A Comprehensive Mathematical Model of Motor Unit Pool Organization, Surface Electromyography, and Force Generation

Neuromuscular physiology is a vibrant research field that has recently seen exciting advances. Previous publications have focused on thorough analyses of particular aspects of neuromuscular physiology, yet an integration of the various novel findings into a single, comprehensive model is missing. In this article, we provide a unified description of a comprehensive mathematical model of surface electromyographic (EMG) measurements and the corresponding force signal in skeletal muscles, both consolidating and extending the results of previous studies regarding various components of the neuromuscular system. The model comprises motor unit (MU) pool organization, recruitment and rate coding, intracellular action potential generation and the resulting EMG measurements, as well as the generated muscular force during voluntary isometric contractions. Mathematically, it consists of a large number of linear PDEs, ODEs, and various stochastic nonlinear relationships, some of which are solved analytically, others numerically. A parameterization of the electrical and mechanical components of the model is proposed that ensures a physiologically meaningful EMG-force relation in the simulated signals, in particular taking the continuous, size-dependent distribution of MU parameters into account. Moreover, a novel nonlinear transformation of the common drive model input is proposed, which ensures that the model force output equals the desired target force. On a physiological level, this corresponds to adjusting the rate coding model to the force generating capabilities of the simulated muscle, while from a control theoretic point of view, this step is equivalent to an exact linearizing transformation of the controlled neuromuscular system. Finally, an alternative analytical formulation of the EMG model is proposed, which renders the physiological meaning of the model more clear and facilitates a mathematical proof that muscle fibers in this model at no point in time represent a net current source or sink. A consistent description of a complete physiological model as presented here, including thorough justification of model component choices, will facilitate the use of these advanced models in future research. Results of a numerical simulation highlight the model's capability to reproduce many physiological effects observed in experimental measurements, and to produce realistic synthetic data that are useful for the validation of signal processing algorithms.

A Scale Mixture-Based Stochastic Model of Surface EMG Signals With Variable Variances

OBJECTIVE

Surface electromyogram (EMG) signals have typically been assumed to follow a Gaussian distribution. However, the presence of non-Gaussian signals associated with muscle activity has been reported in recent studies, and there is no general model of the distribution of EMG signals that can explain both non-Gaussian and Gaussian distributions within a unified scheme.

METHODS

In this paper, we describe the formulation of a non-Gaussian EMG model based on a scale mixture distribution. In the model, an EMG signal at a certain time follows a Gaussian distribution, and its variance is handled as a random variable that follows an inverse gamma distribution. Accordingly, the probability distribution of EMG signals is assumed to be a mixture of Gaussians with the same mean but different variances. The EMG variance distribution is estimated via marginal likelihood maximization.

RESULTS

Experiments involving nine participants revealed that the proposed model provides a better fit to recorded EMG signals than conventional EMG models. It was also shown that variance distribution parameters may reflect underlying motor unit activity.

CONCLUSION

This study proposed a scale mixture distribution-based stochastic EMG model capable of representing changes in non-Gaussianity associated with muscle activity. A series of experiments demonstrated the validity of the model and highlighted the relationship between the variance distribution and muscle force.

SIGNIFICANCE

The proposed model helps to clarify conventional wisdom regarding the probability distribution of surface EMG signals within a unified scheme.

Two-Source Validation of Progressive FastICA Peel-Off for Automatic Surface EMG Decomposition in Human First Dorsal Interosseous Muscle

This study aims to assess the accuracy of a novel high density surface electromyogram (SEMG) decomposition method, namely automatic progressive FastICA peel-off (APFP), for automatic decomposition of experimental electrode array SEMG signals. A two-source method was performed by simultaneous concentric needle EMG and electrode array SEMG recordings from the human first dorsal interosseous (FDI) muscle, using a protocol commonly applied in clinical EMG examination. The electrode array SEMG was automatically decomposed by the APFP while the motor unit action potential (MUAP) trains were also independently identified from the concentric needle EMG. The degree of agreement of the common motor unit (MU) discharge timings decomposed from the two different categories of EMG signals was assessed. A total of 861 and 217 MUs were identified from the 114 trials of simultaneous high density SEMG and concentric needle EMG recordings, respectively. Among them 168 common (MUs) were found with a high average matching rate of [Formula: see text] for the discharge timings. The outcomes of this study show that the APFP can reliably decompose at least a subset of MUs in the high density SEMG signals recorded from the human FDI muscle during low contraction levels using a protocol analog to clinical EMG examination.

One-Channel Surface Electromyography Decomposition for Muscle Force Estimation

Estimating muscle force by surface electromyography (sEMG) is a non-invasive and flexible way to diagnose biomechanical diseases and control assistive devices such as prosthetic hands. To estimate muscle force using sEMG, a supervised method is commonly adopted. This requires simultaneous recording of sEMG signals and muscle force measured by additional devices to tune the variables involved. However, recording the muscle force of the lost limb of an amputee is challenging, and the supervised method has limitations in this regard. Although the unsupervised method does not require muscle force recording, it suffers from low accuracy due to a lack of reference data. To achieve accurate and easy estimation of muscle force by the unsupervised method, we propose a decomposition of one-channel sEMG signals into constituent motor unit action potentials (MUAPs) in two steps: (1) learning an orthogonal basis of sEMG signals through reconstruction independent component analysis; (2) extracting spike-like MUAPs from the basis vectors. Nine healthy subjects were recruited to evaluate the accuracy of the proposed approach in estimating muscle force of the biceps brachii. The results demonstrated that the proposed approach based on decomposed MUAPs explains more than 80% of the muscle force variability recorded at an arbitrary force level, while the conventional amplitude-based approach explains only 62.3% of this variability. With the proposed approach, we were also able to achieve grip force control of a prosthetic hand, which is one of the most important clinical applications of the unsupervised method. Experiments on two trans-radial amputees indicated that the proposed approach improves the performance of the prosthetic hand in grasping everyday objects.

Speedup computation of HD-sEMG signals using a motor unit-specific electrical source model

Nowadays, bio-reliable modeling of muscle contraction is becoming more accurate and complex. This increasing complexity induces a significant increase in computation time which prevents the possibility of using this model in certain applications and studies. Accordingly, the aim of this work is to significantly reduce the computation time of high-density surface electromyogram (HD-sEMG) generation. This will be done through a new model of motor unit (MU)-specific electrical source based on the fibers composing the MU. In order to assess the efficiency of this approach, we computed the normalized root mean square error (NRMSE) between several simulations on single generated MU action potential (MUAP) using the usual fiber electrical sources and the MU-specific electrical source. This NRMSE was computed for five different simulation sets wherein hundreds of MUAPs are generated and summed into HD-sEMG signals. The obtained results display less than 2% error on the generated signals compared to the same signals generated with fiber electrical sources. Moreover, the computation time of the HD-sEMG signal generation model is reduced to about 90% compared to the fiber electrical source model. Using this model with MU electrical sources, we can simulate HD-sEMG signals of a physiological muscle (hundreds of MU) in less than an hour on a classical workstation. Graphical Abstract Overview of the simulation of HD-sEMG signals using the fiber scale and the MU scale. Upscaling the electrical source to the MU scale reduces the computation time by 90% inducing only small deviation of the same simulated HD-sEMG signals.

A new approach for multi-channel surface EMG signal simulation

Simulation models are necessary for testing the performance of newly developed approaches before they can be applied to interpreting experimental data, especially when biomedical signals such as surface electromyogram (SEMG) signals are involved. A new and easily implementable surface EMG simulation model was developed in this study to simulate multi-channel SEMG signals. A single fiber action potential (SFAP) is represented by the sum of three Gaussian functions. SFAP waveforms can be modified by adjusting the amplitude and bandwidth of the Gaussian functions. SEMG signals were successfully simulated at different detected locations. Effects of the fiber depth, electrode position and conduction velocity of SFAP on motor unit action potential (MUAP) were illustrated. Results demonstrate that the easily implementable SEMG simulation approach developed in this study can be used to effectively simulate SEMG signals.

Adipose tissue thickness does not affect the electromechanical delay

During voluntary contractions in humans, the subcutaneous tissues between surface electrodes and active motor units have been shown to attenuate surface electromyographic (EMG) signal amplitude. The purpose of this investigation was to examine the relationship between adipose tissue thickness and the electromechnical delay (EMD) during maximal voluntary contractions (MVCs). Thirty-two healthy women (mean  ±  SD age  =  21  ±  2 years; mass  =  60.7  ±  11.5 kg; height  =  161.7  ±  7.5 cm; dual-energy x-ray absorptiometry body-fat percentage  =  33.1  ±  9.9%) performed MVCs of the right leg extensors while bipolar surface EMG signals were detected from the vastus lateralis muscle. EMD was calculated as the time (ms) between EMG and torque onsets. B-mode ultrasonography was used to determine adipose tissue thickness over the same location of the vastus lateralis where the EMG sensor was placed. Partial correlation was used to examine the relationship between adipose tissue thickness and EMD while statistically removing the influence of peak torque, EMG amplitude, and vastus lateralis muscle thickness. The partial correlation demonstrated no relationship between adipose tissue thickness and EMD (r  =  -0.010, p  =  0.956). Collectively, these findings demonstrated that adiposity does not influence the estimation of EMD.

Planckian Power Spectral Densities from Human Calves during Posture Maintenance and Controlled Isometric Contractions

BACKGROUND

The relationship between muscle anatomy and physiology and its corresponding electromyography activity (EMGA) is complex and not well understood. EMGA models may be broadly divided in stochastic and motor-unit-based models. For example, these models have successfully described many muscle physiological variables such as the value of the muscle fiber velocity and the linear relationship between median frequency and muscle fiber velocity. However they cannot explain the behavior of many of these variables with changes in intramuscular temperature, or muscle PH acidity, for instance. Here, we propose that the motor unit action potential can be treated as an electromagnetic resonant mode confined at thermal equilibrium inside the muscle. The motor units comprising the muscle form a system of standing waves or modes, where the energy of each mode is proportional to its frequency. Therefore, the power spectral density of the EMGA is well described and fit by Planck's law and from its distribution we developed theoretical relationships that explain the behavior of known physiological variables with changes in intramuscular temperature or muscle PH acidity, for instance.

METHODS

EMGA of the calf muscle was recorded during posture maintenance in seven participants and during controlled isometric contractions in two participants. The power spectral density of the EMGA was then fit with the Planckian distribution. Then, we inferred nine theoretical relationships from the distribution and compared the theoretically derived values with experimentally obtained values.

RESULTS

The power spectral density of EMGA was fit by Planckian distributions and all the theoretical relationships were validated by experimental results.

CONCLUSIONS

Only by considering the motor unit action potentials as electromagnetic resonant modes confined at thermal equilibrium inside the muscle suffices to predict known or new theoretical relationships for muscle physiological variables that other models have failed to do.

Innervation zones of fasciculating motor units: observations by a linear electrode array

This study examines the innervation zone (IZ) in the biceps brachii muscle in healthy subjects and those with amyotrophic lateral sclerosis (ALS) using a 20-channel linear electromyogram (EMG) electrode array. Raster plots of individual waveform potentials were studied to estimate the motor unit IZ. While this work mainly focused on fasciculation potentials (FPs), a limited number of motor unit potentials (MUPs) from voluntary activity of 12 healthy and seven ALS subjects were also examined. Abnormal propagation of MUPs and scattered IZs were observed in fasciculating units, compared with voluntarily activated MUPs in healthy and ALS subjects. These findings can be related to muscle fiber reinnervation following motor neuron degeneration in ALS and the different origin sites of FPs compared with voluntary MUPs.

Comparing functional EMG characteristics between zero-order and first-order interface dynamics

The optimal control scheme for powered prostheses can be determined using simulation experiments, for which an accurate model of prosthesis control is essential. This paper focuses on electromyographic (EMG) control signal characteristics across two different control schemes. We constructed a functional EMG model comprising three EMG signal characteristics-standard deviation, kurtosis, and median power frequency-using data collected under realistic conditions for prosthesis control (closed-loop, dynamic, anisometric contractions). We examined how the model changed when subjects used zero-order or first-order control. Control order had a statistically significant effect on EMG characteristics, but the effect size was small and generally did not exceed inter-subject variability. Therefore, we suggest that this functional EMG model remains valid across different control schemes.

A simulation study on the relation between muscle motor unit numbers and the non-Gaussianity/non-linearity levels of surface electromyography

Recent research has demonstrated that surface electromyography (sEMG) signals have non-Gaussianity and non-linearity properties. It is known that more muscle motor units are recruited and firing rates (FRs) increase as exertion increases. A hypothesis was proposed that the Gaussianity test (S (g)) and linearity test (S (ℓ)) levels of sEMG signals are associated with the number of active motor units (nMUs) and the FR. The hypothesis has only been preliminarily discussed in experimental studies. We used a simulation sEMG model involving spatial (active MUs) and temporal (three FRs) information to test the hypothesis. Higher-order statistics (HOS) from the bi-frequency domain were used to perform S (g) and S (ℓ). Multivariate covariance analysis and a correlation test were employed to determine the nMUs-S (g) relationship and the nMUs-S (ℓ) relationship. Results showed that nMUs, the FR, and the interaction of nMUs and the FR all influenced the S (g) and S (ℓ) values. The nMUs negatively correlated to both the S (g) and S (ℓ) values. That is, at the three FRs, sEMG signals tended to a more Gaussian and linear distribution as exertion and nMUs increased. The study limited experiment factors to the sEMG non-Gaussianity and non-linearity levels. The study quantitatively described nMUs and the FR of muscle that are not directly available from experiments. Our finding has guiding significance for muscle capability assessment and prosthetic control.

Methodological aspects of SEMG recordings for force estimation--a tutorial and review

Insight into the magnitude of muscle forces is important in biomechanics research, for example because muscle forces are the main determinants of joint loading. Unfortunately muscle forces cannot be calculated directly and can only be measured using invasive procedures. Therefore, estimates of muscle force based on surface EMG measurements are frequently used. This review discusses the problems associated with surface EMG in muscle force estimation and the solutions that novel methodological developments provide to this problem. First, some basic aspects of muscle activity and EMG are reviewed and related to EMG amplitude estimation. The main methodological issues in EMG amplitude estimation are precision and representativeness. Lack of precision arises directly from the stochastic nature of the EMG signal as the summation of a series of randomly occurring polyphasic motor unit potentials and the resulting random constructive and destructive (phase cancellation) superimpositions. Representativeness is an issue due the structural and functional heterogeneity of muscles. Novel methods, i.e. multi-channel monopolar EMG and high-pass filtering or whitening of conventional bipolar EMG allow substantially less variable estimates of the EMG amplitude and yield better estimates of muscle force by (1) reducing effects of phase cancellation, and (2) adequate representation of the heterogeneous activity of motor units within a muscle. With such methods, highly accurate predictions of force, even of the minute force fluctuations that occur during an isometric and isotonic contraction have been achieved. For dynamic contractions, EMG-based force estimates are confounded by the effects of muscle length and contraction velocity on force producing capacity. These contractions require EMG amplitude estimates to be combined with modeling of muscle contraction dynamics to achieve valid force predictions.

Epoch length to accurately estimate the amplitude of interference EMG is likely the result of unavoidable amplitude cancellation

Researchers and clinicians routinely rely on interference electromyograms (EMGs) to estimate muscle forces and command signals in the neuromuscular system (e.g., amplitude, timing, and frequency content). The amplitude cancellation intrinsic to interference EMG, however, raises important questions about how to optimize these estimates. For example, what should the length of the epoch (time window) be to average an EMG signal to reliably estimate muscle forces and command signals? Shorter epochs are most practical, and significant reductions in epoch have been reported with high-pass filtering and whitening. Given that this processing attenuates power at frequencies of interest (< 250 Hz), however, it is unclear how it improves the extraction of physiologically-relevant information. We examined the influence of amplitude cancellation and high-pass filtering on the epoch necessary to accurately estimate the "true" average EMG amplitude calculated from a 28 s EMG trace (EMG(ref)) during simulated constant isometric conditions. Monte Carlo iterations of a motor-unit model simulating 28 s of surface EMG produced 245 simulations under 2 conditions: with and without amplitude cancellation. For each simulation, we calculated the epoch necessary to generate average full-wave rectified EMG amplitudes that settled within 5% of EMG(ref.) For the no-cancellation EMG, the necessary epochs were short (e.g., < 100 ms). For the more realistic interference EMG (i.e., cancellation condition), epochs shortened dramatically after using high-pass filter cutoffs above 250 Hz, producing epochs short enough to be practical (i.e., < 500 ms). We conclude that the need to use long epochs to accurately estimate EMG amplitude is likely the result of unavoidable amplitude cancellation, which helps to clarify why high-pass filtering (> 250 Hz) improves EMG estimates.

A system for the synchronized recording of sonomyography, electromyography and joint angle

Ultrasound and electromyography (EMG) are two of the most commonly used diagnostic tools for the assessment of muscles. Recently, many studies reported the simultaneous collection of EMG signals and ultrasound images, which were normally amplified and digitized by different devices. However, there is lack of a systematic method to synchronize them and no study has reported the effects of ultrasound gel to the EMG signal collection during the simultaneous data collection. In this paper, we introduced a new method to synchronize ultrasound B-scan images, EMG signals, joint angles and other related signals (e.g. force and velocity signals) in real-time. The B-mode ultrasound images were simultaneously captured by the PC together with the surface EMG (SEMG) and the joint angle signal. The deformations of the forearm muscles induced by wrist motions were extracted from a sequence of ultrasound images, named as Sonomyography (SMG). Preliminary experiments demonstrated that the proposed method could reliably collect the synchronized ultrasound images, SEMG signals and joint angle signals in real-time. In addition, the effect of ultrasound gel on the SEMG signals when the EMG electrodes were close to the ultrasound probe was studied. It was found that the SEMG signals were not significantly affected by the amount of the ultrasound gel. The system is being used for the study of contractions of various muscles as well as the muscle fatigue.

Experimentally valid predictions of muscle force and EMG in models of motor-unit function are most sensitive to neural properties

Computational models of motor-unit populations are the objective implementations of the hypothesized mechanisms by which neural and muscle properties give rise to electromyograms (EMGs) and force. However, the variability/uncertainty of the parameters used in these models--and how they affect predictions--confounds assessing these hypothesized mechanisms. We perform a large-scale computational sensitivity analysis on the state-of-the-art computational model of surface EMG, force, and force variability by combining a comprehensive review of published experimental data with Monte Carlo simulations. To exhaustively explore model performance and robustness, we ran numerous iterative simulations each using a random set of values for nine commonly measured motor neuron and muscle parameters. Parameter values were sampled across their reported experimental ranges. Convergence after 439 simulations found that only 3 simulations met our two fitness criteria: approximating the well-established experimental relations for the scaling of EMG amplitude and force variability with mean force. An additional 424 simulations preferentially sampling the neighborhood of those 3 valid simulations converged to reveal 65 additional sets of parameter values for which the model predictions approximate the experimentally known relations. We find the model is not sensitive to muscle properties but very sensitive to several motor neuron properties--especially peak discharge rates and recruitment ranges. Therefore to advance our understanding of EMG and muscle force, it is critical to evaluate the hypothesized neural mechanisms as implemented in today's state-of-the-art models of motor unit function. We discuss experimental and analytical avenues to do so as well as new features that may be added in future implementations of motor-unit models to improve their experimental validity.

Development of a Synchronized System for Continuous Acquisition and Analysis of Ultrasound Joint Angle, and EMG

Ultrasound and electromyography (EMG) are two of the most often used diagnostic tools for muscles. However, there is lack of a system for continuously capturing both signals with a synchronized procedure. This study presents a new system for this purpose. The system comprises an ultrasound scanner, a pulser/receiver, an EMG amplifier, and a PC with A/D converter cards for data acquisition and analysis. The A-mode and B-mode ultrasound data, which were digitized by the A/D converter and a video capture card respectively, could be simultaneously captured by the PC together with the surface EMG (SEMG) signal, which was digitized by another data acquisition card. Time markers for all frames of ultrasound and SEMG signals were recorded to synchronize the two data streams. The tissue deformation was extracted inform the ultrasound signals or images using cross-correlation algorithms. The RMS and spectrum of SEMG was calculated to study the muscle activity. In addition, the joint angle signal was also be synchronized with ultrasound and SEMG signals. The experiments involving 5 subjects were carried out to test the performance of the system. Preliminary results showed that the muscle deformations extracted from the ultrasound data were well correlated with the changes of SEMG RMS. The system may have potential values in the investigation of muscle properties and activities.

Clinical applications of high-density surface EMG: A systematic review

High density-surface EMG (HD-sEMG) is a non-invasive technique to measure electrical muscle activity with multiple (more than two) closely spaced electrodes overlying a restricted area of the skin. Besides temporal activity HD-sEMG also allows spatial EMG activity to be recorded, thus expanding the possibilities to detect new muscle characteristics. Especially muscle fiber conduction velocity (MFCV) measurements and the evaluation of single motor unit (MU) characteristics come into view. This systematic review of the literature evaluates the clinical applications of HD-sEMG. Although beyond the scope of the present review, the search yielded a large number of "non-clinical" papers demonstrating that a considerable amount of work has been done and that significant technical progress has been made concerning the feasibility and optimization of HD-sEMG techniques. Twenty-nine clinical studies and four reviews of clinical applications of HD-sEMG were considered. The clinical studies concerned muscle fatigue, motor neuron diseases (MND), neuropathies, myopathies (mainly in patients with channelopathies), spontaneous muscle activity and MU firing rates. In principle, HD-sEMG allows pathological changes at the MU level to be detected, especially changes in neurogenic disorders and channelopathies. We additionally discuss several bioengineering aspects and future clinical applications of the technique and provide recommendations for further development and implementation of HD-sEMG as a clinical diagnostic tool.

An analytical model for surface EMG generation in volume conductors with smooth conductivity variations

A nonspace invariant model of volume conductor for surface electromyography (EMG) signal generation is analytically investigated. The volume conductor comprises planar layers representing the muscle and subcutaneous tissues. The muscle tissue is homogeneous and anisotropic while the subcutaneous layer is inhomogeneous and isotropic. The inhomogeneity is modeled as a smooth variation in conductivity along the muscle fiber direction. This may reflect a practical situation of tissues with different conductivity properties in different locations or of transitions between tissues with different properties. The problem is studied with the regular perturbation theory, through a series expansion of the electric potential. This leads to a set of Poisson's problems, for which the source term in an equation and the boundary conditions are determined by the solution of the previous equations. This set of problems can be solved iteratively. The solution is obtained in the two-dimensional Fourier domain, with spatial angular frequencies corresponding to the longitudinal and perpendicular direction with respect to the muscle fibers, in planes parallel to the detection surface. The series expansion is truncated for the practical implementation. Representative simulations are presented. The proposed model constitutes a new approach for surface EMG signal simulation with applications related to the validation of methods for information extraction from this signal.