Surface Electromyography Features Research Articles

Poster (Technology Innovation) ID 1984794

Background Cervical spinal cord injury (SCI) can cause significant impairment and disability with an impact on individuals’ quality of life and independence. Surface electromyography (SEMG) is a sensitive and non-invasive technique to measure muscle activity and has demonstrated great potential in capturing the impact from SCI. The mechanisms of SCI damage on SEMG signal characteristics are multi-faceted and difficult to study in vivo. Objective Use validated computational models to characterize changes in SEMG signal after SCI and identify SEMG features that are sensitive and specific to the impact from different aspects of SCI. Method Starting from existing computational models for motor neuron pool organization and for motor unit action potential generation for healthy neuromuscular systems, we set up scenarios to model alterations in upper motor neurons, lower motor neurons, and the number of muscle fibers within each motor unit after SCI. After simulating SEMG signals from each scenario, we extracted time and frequency domain features and investigated the impact of SCI disruptions on SEMG features using the Pearson correlation between a feature and the extent of a given disruption. Findings Commonly used amplitude-based SEMG features cannot differentiate between injury scenarios. A broader set of features provides greater specificity to the type of damage present. Conclusion We demonstrated a novel approach to mechanistically relate SEMG features to different types of neuromuscular alterations after SCI. This work contributes to a deeper understanding and exploitation of SEMG in clinical applications, which will ultimately improve patient outcomes after SCI.

Evaluating Electromyography and Sonomyography Sensor Fusion to Estimate Lower-Limb Kinematics Using Gaussian Process Regression

Research on robotic lower-limb assistive devices over the past decade has generated autonomous, multiple degree-of-freedom devices to augment human performance during a variety of scenarios. However, the increase in capabilities of these devices is met with an increase in the complexity of the overall control problem and requirement for an accurate and robust sensing modality for intent recognition. Due to its ability to precede changes in motion, surface electromyography (EMG) is widely studied as a peripheral sensing modality for capturing features of muscle activity as an input for control of powered assistive devices. In order to capture features that contribute to muscle contraction and joint motion beyond muscle activity of superficial muscles, researchers have introduced sonomyography, or real-time dynamic ultrasound imaging of skeletal muscle. However, the ability of these sonomyography features to continuously predict multiple lower-limb joint kinematics during widely varying ambulation tasks, and their potential as an input for powered multiple degree-of-freedom lower-limb assistive devices is unknown. The objective of this research is to evaluate surface EMG and sonomyography, as well as the fusion of features from both sensing modalities, as inputs to Gaussian process regression models for the continuous estimation of hip, knee and ankle angle and velocity during level walking, stair ascent/descent and ramp ascent/descent ambulation. Gaussian process regression is a Bayesian nonlinear regression model that has been introduced as an alternative to musculoskeletal model-based techniques. In this study, time-intensity features of sonomyography on both the anterior and posterior thigh along with time-domain features of surface EMG from eight muscles on the lower-limb were used to train and test subject-dependent and task-invariant Gaussian process regression models for the continuous estimation of hip, knee and ankle motion. Overall, anterior sonomyography sensor fusion with surface EMG significantly improved estimation of hip, knee and ankle motion for all ambulation tasks (level ground, stair and ramp ambulation) in comparison to surface EMG alone. Additionally, anterior sonomyography alone significantly improved errors at the hip and knee for most tasks compared to surface EMG. These findings help inform the implementation and integration of volitional control strategies for robotic assistive technologies.

Feature Extraction and Evaluation for Classification Models of Injurious Falls Based on Surface Electromyography

Background: Only 2% of falls in older adults result in serious injuries (i.e., hip fracture). Therefore, it is important to differentiate injurious versus non-injurious falls, which is critical to develop effective interventions for injury prevention. Objects: The purpose of this study was to a. extract the best features of surface electromyography (sEMG) for classification of injurious falls, and b. find a best model provided by data mining techniques using the extracted features. Methods: Twenty young adults self-initiated falls and landed sideways. Falling trials were consisted of three initial fall directions (forward, sideways, or backward) and three knee positions at the time of hip impact (the impacting-side knee contacted the other knee (“knee together”) or the mat (“knee on mat”), or neither the other knee nor the mat was contacted by the impacting-side knee (“free knee”). Falls involved “backward initial fall direction” or “free knee” were defined as “injurious falls” as suggested from previous studies. Nine features were extracted from sEMG signals of four hip muscles during a fall, including integral of absolute value (IAV), Wilson amplitude (WAMP), zero crossing (ZC), number of turns (NT), mean of amplitude (MA), root mean square (RMS), average amplitude change (AAC), difference absolute standard deviation value (DASDV). The decision tree and support vector machine (SVM) were used to classify the injurious falls. Results: For the initial fall direction, accuracy of the best model (SVM with a DASDV) was 48%. For the knee position, accuracy of the best model (SVM with an AAC) was 49%. Furthermore, there was no model that has sensitivity and specificity of 80% or greater. Conclusion: Our results suggest that the classification model built upon the sEMG features of the four hip muscles are not effective to classify injurious falls. Future studies should consider other data mining techniques with different muscles.

SEMG Feature Extraction Based on StockwellTransform Improves Hand MovementRecognition Accuracy.

Feature extraction, as an important method for extracting useful information from surface electromyography (SEMG), can significantly improve pattern recognition accuracy. Time and frequency analysis methods have been widely used for feature extraction, but these methods analyze SEMG signals only from the time or frequency domain. Recent studies have shown that feature extraction based on time-frequency analysis methods can extract more useful information from SEMG signals. This paper proposes a novel time-frequency analysis method based on the Stockwell transform (S-transform) to improve hand movement recognition accuracy from forearm SEMG signals. First, the time-frequency analysis method, S-transform, is used for extracting a feature vector from forearm SEMG signals. Second, to reduce the amount of calculations and improve the running speed of the classifier, principal component analysis (PCA) is used for dimensionality reduction of the feature vector. Finally, an artificial neural network (ANN)-based multilayer perceptron (MLP) is used for recognizing hand movements. Experimental results show that the proposed feature extraction based on the S-transform analysis method can improve the class separability and hand movement recognition accuracy compared with wavelet transform and power spectral density methods.

Beta-band motor unit coherence and nonlinear surface EMG features of the first dorsal interosseous muscle vary with force.

Motor unit firing times are weakly coupled across a range of frequencies during voluntary contractions. Coherent activity within the beta-band (15-35 Hz) has been linked to oscillatory cortical processes, providing evidence of functional connectivity between the motoneuron pool and motor cortex. The aim of this study was to investigate whether beta-band motor unit coherence is altered with increasing abduction force in the first dorsal interosseous muscle. Coherence between motor unit firing times, extracted from decomposed surface electromyography (EMG) signals, was investigated in 17 subjects at 10, 20, 30, and 40% of maximum voluntary contraction. Corresponding changes in nonlinear surface EMG features (specifically sample entropy and determinism, which are sensitive to motor unit synchronization) were also examined. A reduction in beta-band and alpha-band coherence was observed as force increased [F(3, 151) = 32, P < 0.001 and F(3, 151) = 27, P < 0.001, respectively], accompanied by corresponding changes in nonlinear surface EMG features. A significant relationship between the nonlinear features and motor unit coherence was also detected (r = -0.43 ± 0.1 and r = 0.45 ± 0.1 for sample entropy and determinism, respectively; both P < 0.001). The reduction in beta-band coherence suggests a change in the relative contribution of correlated and uncorrelated presynaptic inputs to the motoneuron pool, and/or a decrease in the responsiveness of the motoneuron pool to synchronous inputs at higher forces. The study highlights the importance of considering muscle activation when investigating changes in motor unit coherence or nonlinear EMG features and examines other factors that can influence coherence estimation.NEW & NOTEWORTHY Intramuscular alpha- and beta-band coherence decreased as muscle contraction force increased. Beta-band coherence was higher in groups of high-threshold motor units than in simultaneously active lower threshold units. Alterations in motor unit coherence with increases or decreases in force and with the onset of fatigue were accompanied by corresponding changes in surface electromyography sample entropy and determinism. Mixed-model analysis indicated mean firing rate and number of motor units also influenced the coherence estimate.

Virtual Rehabilitation Training System Based on Surface EMG Feature Extraction and Analysis.

Aiming at the characteristics that electromyography (EMG) signals can reflect the human body's motive intention and the information of muscle's motive state, this paper makes a thorough study on the evaluation of surface electromyography signals' motive state. At the same time, EMG signals can reflect the characteristics of limb movement and its changing rules, and can acquire the functional characteristics of limb movement so as to accurately evaluate the rehabilitation status of patients. In this paper, EMG signal analysis and feedback control are introduced into the virtual rehabilitation system to study the methods of EMG parameter identification and dynamic feature extraction, and obtain the EMG characteristics and variation rules related to human motion patterns. In this paper, a rehabilitation training system based on EMG feedback and virtual reality is built, and the validity of the system is verified by patient experiment. The feasibility of the system is verified by the methods of validity of the algorithm, recognition rate of the system action pattern and fatigue evaluation.

Increased EMG intermuscular coherence and reduced signal complexity in Parkinson’s disease

ObjectivesTo investigate differences in surface electromyography (EMG) features in individuals with idiopathic Parkinson’s disease (PD) and aged-matched controls. MethodsSurface EMG was recorded during isometric leg extension in PD patients prior to, and after undergoing a locomotor training programme, and in aged-matched controls. Differences in EMG structure were quantified using determinism (%DET), sample entropy (SampEn) and intermuscular coherence. Results%DET was significantly higher, and SampEn significantly lower, in PD patients. Intermuscular coherence was also significantly higher in the PD group in theta, alpha and beta frequency bands. %DET increased and SampEn decreased with increasing Movement-Disorder-Society UPDRS scores, while theta band coherence was significantly correlated with total MDS-UPDRS scores and torque variance. Neither %DET, SampEn nor intermuscular coherence changed in response to training. ConclusionsThe differences observed are consistent with increased synchrony among motor units within and across leg muscles in PD. Differences between EMG signals recorded from the PD and control groups persisted post-therapy, after improvements in walking capacity occurred. SignificanceThese results provide insight into changes in motoneuron activity in PD, demonstrate increased beta band intramuscular coherence in PD for the first time, and support the development of quantitative biomarkers for PD based on advanced surface EMG features.

Surface EMG based continuous estimation of human lower limb joint angles by using deep belief networks

Abstract Surface electromyography (EMG) signals have been widely used in locomotion studies and human-machine interface applications. In this paper, a regression model which relates the multichannel surface EMG signals to human lower limb flexion/extension (FE) joint angles is constructed. In the experimental paradigm, three dimensional trajectories of 16 external markers on the human lower limbs were recorded by optical motion capture system and surface EMG signals from 10 muscles directly concerned with the lower limb motion were recorded synchronously. With the raw data, the joint angles of hip, knee and ankle were calculated accurately and the time series of intensity for surface EMG signals were extracted. Then, a deep belief networks (DBN) that consists of restricted Boltzmann machines (RBM) was built, by which the multi-channel processed surface EMG signals were encoded in low dimensional space and the optimal features were extracted. Finally, a back propagation (BP) neural network was used to map the optimal surface EMG features to the FE joint angles. The results show that, the features extracted from multichannel surface EMG signals using DBN method proposed in this paper outperform principal components analysis (PCA), and the root mean square error (RMSE) between the estimated joint angles and calculated ones during human walking is reduced by about 50%. The proposed model is expected to develop human-machine interaction interface to achieve continuous bioelectric control and to improve motion stability between human and machine, especially for lower limb wearable intelligent equipment.

Spectral and mathematical evaluation of electromyography signals for clinical use

The surface electromyography (SEMG) is a complicated biomedical signal, generated during voluntary or involuntary muscle activities and these muscle activities are always controlled by the nervous system. In this paper, the processing and analysis of SEMG signals at multiple muscle points for different operations were carried out. Myoelectric signals were detected using designed acquisition setup which consists of an instrumentation amplifier, filter circuit, an amplifier with gain adjustment. Further, Labview[Formula: see text]-based data programming code was used to record SEMG signals for independent activities. The whole system consists of bipolar noninvasive electrodes, signal acquisition protocols and signal conditioning at different levels. This work uses recorded SEMG signals generated by biceps and triceps muscles for four different arm activities. Feature extraction was done on the recorded signal for investigating the voluntary muscular contraction relationship for exercising statistic measured index method to evaluate distance between two independent groups by directly addressing the quality of signal in separability class for different arm movements. Thereafter repeated factorial analysis of variance technique was implemented to evaluate the effectiveness of processed signal. From these results, it demonstrates that the proposed method can be used as SEMG feature evaluation index.

Joint force estimation using time-varying SEMG feature in fatiguing contraction.

Many studies have estimated joint force using surface electromyography (SEMG), however, the time-variant characteristic of SEMG is not considered. The change of SEMG amplitude is one of manifestations of muscle fatigue. This study proposes a force estimation method using SEMG in fatiguing contraction. The SEMG amplitude is used to determine the signal states by k-means clustering method. According to the signal state changes, the corresponding gain is used to estimate the force. The target contraction is an isometric abduction of an index finger in static and dynamic force conditions for 5 healthy subjects. The estimation performance was evaluated by percentage of root mean squared error (RMSE). The RMSE for the proposed method is 2.5 ± 1.0% under static condition and 8.8 ± 1.2% under dynamic condition. The accuracy using a constant gain calculated at initial time was used to compare with the proposed method. The RMSE are 8.9 ± 2.2% under static condition and 10.1 ± 2.4% under dynamic condition. The proposed method had better performance in both conditions.

Quantitative analysis of surface electromyography during epileptic and nonepileptic convulsive seizures.

To investigate the characteristics of sustained muscle activation during convulsive epileptic and psychogenic nonepileptic seizures (PNES), as compared to voluntary muscle activation. The main goal was to find surface electromyography (EMG) features that can distinguish between convulsive epileptic seizures and convulsive PNES. In this case-control study, surface EMG was recorded from the deltoid muscles during long-term video-electroencephalography (EEG) monitoring in 25 patients and in 21 healthy controls. A total of 46 clinical episodes were recorded: 28 generalized tonic-clonic seizures (GTCS) from 14 patients with epilepsy, and 18 convulsive PNES from 12 patients (one patient had both GTCS and PNES). The healthy controls were simulating GTCS. To quantitatively characterize the signals we calculated the following parameters: root mean square (RMS) of the amplitude, median frequency (MF), coherence, and duration of the seizures, of the clonic EMG discharges, and of the silent periods between the cloni. Based on wavelet analysis, we distinguished between a low-frequency component (LF 2-8 Hz) and a high-frequency component (HF 64-256 Hz). Duration of the seizure, and separation between the tonic and the clonic phases distinguished at group-level but not at individual level between convulsive PNES and GTCS. RMS, temporal dynamics of the HF/LF ratio, and the evolution of the silent periods differentiated between epileptic and nonepileptic convulsive seizures at the individual level. A combination between HF/LF ratio and RMS separated all PNES from the GTCS. A blinded review of the EMG features distinguished correctly between GTCS and convulsive PNES in all cases. The HF/LF ratio and the RMS of the PNES were smaller compared to the simulated seizures. In addition to providing insight into the mechanism of muscle activation during convulsive PNES, these results have diagnostic significance, at the individual level. Surface EMG features can accurately distinguish convulsive epileptic from nonepileptic psychogenic seizures, even in PNES cases without rhythmic clonic movements.

Multi-Scale Surface Electromyography Modeling to Identify Changes in Neuromuscular Activation With Myofascial Pain

To solve the limitations in using the conventional parametric measures to define myofascial pain, a 3-D multi-scale wavelet energy variation graph is proposed as a way to inspect the pattern of surface electromyography (SEMG) variation between the dominant and nondominant sides at different frequency scales during a muscle contraction cycle and the associated changes with the upper-back myofascial pain. The model was developed based on the property of the wavelet energy of the SEMG signal revealing the degree of correspondence between the shape of the motor unit action potential and the wavelet waveform at a certain scale in terms of the frequency band. The characteristic pattern of the graph for each group (30 normal and 26 patient subjects) was first derived and revealed the dominant-hand effect and the changes with myofascial pain. Through comparison of individual graphs across subjects, we found that the graph pattern reveals a sensitivity of 53.85% at a specificity of 83.33% in the identification of myofascial pain. The changes in these patterns provide insight into the transformation between different fiber recruitment, which cannot be explored using conventional SEMG features. Therefore, this multi-scale analysis model could provide a reliable SEMG features to identify myofascial pain.

Quantitative and sensitive assessment of neurophysiological status after human spinal cord injury

This study was designed to develop an objective and sensitive spinal cord injury (SCI) characterization protocol based on surface electromyography (EMG) activity. Twenty-four patients at both acute and chronic time points post-SCI, as well as 4 noninjured volunteers, were assessed using neurophysiological and clinical measures of volitional motor function. The EMG amplitude was recorded from 15 representative muscles bilaterally during standardized maneuvers as a neurophysiological assessment of voluntary motor function. International Standards for the Neurological Classification of Spinal Cord Injury (ISNCSCI) examinations were performed as a clinical assessment of lesion severity. Sixty-six functional neurophysiological assessments were performed in 24 patients with SCI and in 4 neurologically intact individuals. The collected EMG data were organized by quantitative parameters and statistically analyzed. The correlation between root mean square (RMS) of the EMG signals and ISNCSCI motor score was confirmed by Kendall correlation analysis. The Kendall correlation value between overall muscles/levels, motor scores, and the RMS of the EMG data is 0.85, with the 95% CI falling into the range of 0.76-0.95. Significant correlations were also observed for the soleus (0.51 [0.28-0.74]), tibialis anterior (TA) (0.53 [0.33-0.73]), tricep (0.52, [0.34-0.70]), and extensor carpi radialis (ECR) (0.80 [0.42-1.00]) muscles. Comparisons of RMS EMG values in groups defined by ISNCSCI motor score further confirmed these results. At the bicep and ECR, patients with motor scores of 5 had nearly significantly higher RMS EMG values than patients with motor scores of 0 (p = 0.059 and 0.052, respectively). At the soleus and TA, the RMS of the EMG value was significantly higher (p < 0.01) for patients with American Spinal Injury Association Impairment Scale motor scores of 5 than for those with ISNCSCI motor scores of 0. Those with C-7 ISNCSCI motor scores of 5 had significantly higher RMS EMG values at the tricep than those with motor scores of 4 (p = 0.008) and 0 (p = 0.02). Results also show that surface EMG signals recorded from trunk muscles allowed the examiner to pick up subclinical changes, even though no ISNCSCI scores were given. Surface EMG signal is suitable for objective neurological SCI characterization protocol design. The quantifiable features of surface EMG may increase SCI characterization resolution by adding subclinical details to the clinical picture of lesion severity and distribution.

Surface electromyography features in manual workers affected by carpal tunnel syndrome

Alterations in surface electromyographic (sEMG) signals of the abductor pollicis brevis muscle were evaluated in 24 non-manual workers and 40 manual workers (25 asymptomatic and 15 reporting CTS symptoms). The initial value (IV) and the normalized rate of change (NRC) of average rectified value (ARV), mean frequency of the power spectrum (MNF), and muscle fiber conduction velocity (MFCV) were calculated during contractions at 20% and 50% of maximal voluntary contraction (MVC). Neuromuscular efficiency (NME) and kurtosis of the sEMG amplitude distribution were estimated. With respect to controls, manual workers showed higher NME, lower ARV IV, and reduced myoelectric manifestations of fatigue (lower MNF NRC for both contraction levels, and lower MFCV NRC at 50% MVC). Kurtosis at 20% MVC showed higher values in symptomatic manual workers than in the other two groups. Kurtosis seems to be a promising parameter for use in monitoring individuals who develop CTS.

Lumbar muscle electromyographic dynamic topography during flexion-extension

The objective of this study is to introduce dynamic topography of surface electromyography (SEMG) to visualize lumbar muscle myoelectric activity and provides a new view to analyze muscle activity in vivo. A total of 20 healthy male subjects and 15 males LBP were enrolled. An electrode-array was applied to the lumbar region to collect SEMG. The root mean square (RMS) value was calculated for each channel, and then a 160×120 matrix was constructed using a linear cubic spline interpolation of each scan to create a 2-D color topographic image. Along a definite interval of action, a series of RMS topography matrices was concatenated as a function of position and time, to form a dynamic topographical video of lumbar muscle activity. Relative area (RA), relative width (RW), relative height (RH) and Width-to-Height Ratio ( W/ H) were chosen as the four quantitative parameters in measuring topographic features. Normal RMS dynamic topography was found to have a consistent, symmetric pattern with a high intensity area in the paraspinal area. LBP patients had a different RMS dynamic topography, with an asymmetric, broad, or disorganized distribution. Quantitative SEMG features were found significantly different between normal control and LBP. After physiotherapy rehabilitation, the dynamic topography images of LBP tended towards the normal pattern. There are obvious differences in lumbar muscle coordination between healthy subjects and LBP patients. The dynamic topography allows the continuous visualization of the distribution of surface EMG signals and the coordination of muscular contractions.

CHARACTERIZATION OF FINGER ISOMETRIC FORCE PRODUCTION WITH MAXIMUM POWER OF SURFACE ELECTROMYOGRAPHY

Finger's action has been controlled by both intrinsic and extrinsic hand muscles. Characterizing the finger action with the activations of hand muscles could be useful for evaluating the neuromuscular control strategy of finger's motor functions. This study is designed to explore the correlation of isometric fingertip force production and frequency-domain features of surface electromyography (sEMG) recorded on extrinsic hand muscles. To this end, 13 subjects (five male and eight female university students) have been recruited to conduct a target force-tracking task. Each subject is required to produce a certain level of force with either the index or middle fingertip to match the pseudo-random ordered target force level (4N, 6N, or 8N) as accurate as possible. During the finger force production process, the sEMG signals are recorded on two extrinsic hand muscles: flex digitorum superficials (FDS) and extensor digitorum (ED). For each sEMG trail, the power spectrum is estimated with the autoregressive (AR) model and from which the maximum power is obtained. Our experimental results reveal three findings: (1) the maximum power increases with the force level regardless of the force producing finger (i.e. index or middle) and the extrinsic hand muscle (i.e. FDS or ED). (2) The sEMG maximum power of index finger is significantly lower than that of the middle finger under the same force level and extrinsic hand muscle. (3) No significant difference can be found between the maximum powers of FDS and ED. The results indicate that the activations of the extrinsic muscles are affected by both the force level and the force producing finger. Based on our findings, the sEMG maximum power of the extrinsic hand muscles could be used as a key parameter to describe the finger's actions.

Modeling of surface myoelectric signals--Part I: Model implementation.

The relationships between the parameters of active motor units (MU's) and the features of surface electromyography (EMG) signals have been investigated using a mathematical model that represents the surface EMG as a summation of contributions from the single muscle fibers. Each MU has parallel fibers uniformly scattered within a cylindrical volume of specified radius embedded in an anisotropic medium. Two action potentials, each modeled as a current tripole, are generated at the neuromuscular junction, propagate in opposite directions and extinguish at the fiber-tendon endings. The neuromuscular junctions and fiber-tendon endings are uniformly scattered within regions of specified width. Muscle fiber conduction velocity and average fiber length to the right and left of the center of the innervation zone are also specified. The signal produced by MU's with different geometries and conduction velocities are superimposed. Monopolar, single differential and double differential signals are computed from electrodes placed in equally spaced locations on the surface of the muscle and are displayed as functions of any of the model's parameters. Spectral and amplitude variables and conduction velocity are estimated from the surface signals and displayed as functions of any of the model's parameters. The influence of fiber-end effects, electrode misalignment, tissue anisotropy, MU's location and geometry are discussed. Part II of this paper will focus on the simulation and interpretation of experimental signals.