Simulated Electromyographic Research Articles

Tutorial on MUedit: An open-source software for identifying and analysing the discharge timing of motor units from electromyographic signals

We introduce the open-source software MUedit and we describe its use for identifying the discharge timing of motor units from all types of electromyographic (EMG) signals recorded with multi-channel systems. MUedit performs EMG decomposition using a blind-source separation approach. Following this, users can display the estimated motor unit pulse trains and inspect the accuracy of the automatic detection of discharge times. When necessary, users can correct the automatic detection of discharge times and recalculate the motor unit pulse train with an updated separation vector. Here, we provide an open-source software and a tutorial that guides the user through (i) the parameters and steps of the decomposition algorithm, and (ii) the manual editing of motor unit pulse trains. Further, we provide simulated and experimental EMG signals recorded with grids of surface electrodes and intramuscular electrode arrays to benchmark the performance of MUedit. Finally, we discuss advantages and limitations of the blind-source separation approach for the study of motor unit behaviour during tonic muscle contractions.

2CFastICA: A Novel Method for High Density Surface EMG Decomposition Based on Kernel Constrained FastICA and Correlation Constrained FastICA.

This study presents a novel high density surface electromyography (EMG) decomposition method, named as 2CFastICA, because it incorporates two key algorithms: kernel constrained FastICA and correlation constrained FastICA. The former focuses on overcoming the local convergence of FastICA without requiring the peel-off strategy used in the progressive FastICA peel-off (PFP) framework. The latter further refines the output of kernel constrained FastICA by correcting possible erroneous or missed spikes. The two constrained FastICA algorithms supplement each other to warrant the decomposition performance. The 2CFastICA method was validated using simulated surface EMG signals with different motor unit numbers and signal to noise ratios (SNRs). Two source validation was also performed by simultaneous high density surface EMG and intramuscular EMG recordings, showing a matching rate (MR) of (97.2 ± 3.5)% for 170 common motor units. In addition, a different form of two source validation was also conducted taking advantages of the high density surface EMG characteristics of patients with amyotrophic lateral sclerosis, showing a MR of (99.4 ± 0.9)% for 34 common motor units from interference and sparse datasets. Both simulation and experimental results indicate that 2CFastICA can achieve similar decomposition performance to PFP. However, the efficiency of decomposition can be greatly improved by 2CFastICA since the complex signal processing procedures associated with the peel-off strategy are not required any more. Along with this paper, we also provide the MATLAB open source code of 2CFastICA for high density surface EMG decomposition.

Assessment of the Non-linear Response of the fSampEn on Simulated EMG Signals.

Fixed sample entropy (fSampEn) is a promising technique for the analysis of respiratory electromyographic (EMG) signals. Its use has shown outperformance of amplitude-based estimators such as the root mean square (RMS) in the evaluation of respiratory EMG signals with cardiac noise and a high correlation with respiratory signals, allowing changes in respiratory muscle activity to be tracked. However, the relationship between the fSampEn response to a given muscle activation has not been investigated. The aim of this study was to analyze the nature of the fSampEn measurements that are produced as the EMG activity increases linearly. Simulated EMG signals were generated and increased linearly. The effect of the parameters r and the size of the moving window N of the fSampEn were evaluated and compared with those obtained using the RMS. The RMS showed a linear trend throughout the study. A non-linear, sigmoidal-like behavior was found when analyzing the EMG signals using the fSampEn. The lower the values of r, the higher the non-linearity observed in the fSampEn results. Greater moving windows reduced the variation produced by too small values of r.Clinical Relevance- Understanding the inherent non-linear relationship produced when using the fSampEn in EMG recordings will contribute to the improvement of the respiratory muscle activation assessment at different levels of respiratory effort in patients with respiratory conditions, particularly during the inspiratory phase.

Adaptive Real-Time Identification of Motor Unit Discharges From Non-Stationary High-Density Surface Electromyographic Signals.

Estimation of the discharge pattern of motor units by electromyography (EMG) decomposition has been applied for neurophysiologic investigations, clinical diagnosis, and human-machine interfacing. However, most of the methods for EMG decomposition are currently applied offline. Here, we propose an approach for high-density surface EMG decomposition in real-time. A real-time decomposition scheme including two sessions, offline training and online decomposition, is proposed based on the convolutional kernel compensation algorithm. The estimation parameters, separation vectors and the thresholds for spike extraction, are first computed during offline training, and then they are directly applied to estimate motor unit spike trains (MUSTs) during the online decomposition. The estimation parameters are updated with the identification of new discharges to adapt to non-stationary conditions. The decomposition accuracy was validated on simulated EMG signals by convolving synthetic MUSTs with motor unit action potentials (MUAPs). Moreover, the accuracy of the online decomposition was assessed from experimental signals recorded from forearm muscles using a signal-based performance metrics (pulse-to-noise ratio, PNR). The proposed algorithm yielded a high decomposition accuracy and robustness to non-stationary conditions. The accuracy of MUSTs identified from simulated EMG signals was 80% for most conditions. From experimental EMG signals, on average, 12 ± 2 MUSTs were identified from each electrode grid with PNR of 25.0 ± 1.8dB, corresponding to an estimated decomposition accuracy 75%. These results indicate the feasibility of real-time identification of motor unit activities non-invasively during variable force contractions, extending the potential applications of high-density EMG as a neural interface.

Can Wavelet Denoising Improve Motor Unit Potential Template Estimation?

Background: Electromyographic (EMG) signals obtained from a contracted muscle contain valuable information on its activity and health status. Much of this information lies in motor unit potentials (MUPs) of its motor units (MUs), collected during the muscle contraction. Hence, accurate estimation of a MUP template for each MU is crucial.Objective: To investigate the possibility of improving MUP template estimation using the wavelet denoising technique.Material and Methods: In this analytical study, several MUP template estimators were developed by combining conventional estimation methods and wavelet denoising techniques. A MUP template was initially estimated using conventional methods such as mean, median, median-trimmed mean, or mode. Thereafter, it was post-processed using the wavelet denoising technique. Results: Evaluation results of the studied estimators using 40 simulated EMG signals with a true template for each constituent MUP train showed that augmented wavelet- based template estimation methods are more reliable than conventional methods. However, on average, wavelet denoising was not much effective. Around 40 MUPs of a MU is sufficient to estimate its MUP template.Conclusions: Although wavelet techniques are effective in EMG signal analysis, here wavelet denoising did not practically improve MUP template estimation. Considering computational simplicity and estimation error, the two methods median and median-trimmed mean are practical estimators that can provide a good estimation of a MUP template for a MU when approximately 40 MUPs are available. Nevertheless, the baseline noise level in the MUP templates estimated using the median-trimmed mean method is slightly lower than that in the templates estimated using the median method.

Signal-to-Noise Ratio Estimation in Electromyography Signals Contaminated with Electrocardiography Signals

When electromyography (EMG) signals are collected from muscles in the torso, they can be perturbed by the electrocardiography (ECG) signals from heart activity. In this paper, we present a novel signal-to-noise ratio (SNR) estimate for an EMG signal contaminated by an ECG signal. We use six features that are popular in assessing EMG signals, namely skewness, kurtosis, mean average value, waveform length, zero crossing and mean frequency. The features were calculated from the raw EMG signals and the detail coefficients of the discrete stationary wavelet transform. Then, these features are used as inputs to a neural network that outputs the estimate of SNR. While we used simulated EMG signals artificially contaminated with simulated ECG signals as the training data, the testing was done with simulated EMG signals artificially contaminated with real ECG signals. The results showed that the waveform length determined with raw EMG signals was the best feature for estimating SNR. It gave the highest average correlation coefficient of 0.9663. These results suggest that the waveform length could be deployed not only in EMG recognition systems but also in EMG signal quality measurements when the EMG signals are contaminated by ECG interference.

Inverse dynamic estimates of muscle recruitment and joint contact forces are more realistic when minimizing muscle activity rather than metabolic energy or contact forces

BackgroundAssessment of contact forces is essential for a better understanding of mechanical factors affecting progression of osteoarthritis. Since contact forces cannot be measured non-invasively, computer simulations are often used to assess joint loading. Contact forces are to a large extent determined by muscle forces. These muscle forces are computed using optimization techniques that solve the muscle redundancy problem by assuming that muscles are coordinated in a way that optimizes performance (e.g., minimizes muscle activity or metabolic energy). However, it is unclear which of the many proposed performance criteria best describes muscle coordination. Research questionWhich performance criterion best describes muscle recruitment patterns and knee contact forces recorded using electromyography (EMG) and load cell instrumented prostheses?. MethodsWe solved the muscle redundancy problem based on six different groups of performance criteria: muscle activations, volume-scaled activations, forces, stresses, metabolic energy, and joint contact forces. Computed muscle excitations and knee contact forces during over-ground walking were validated against recorded EMG signals and measured contact forces for four subjects with instrumented knee prostheses in the “Grand Challenge Competition to Predict in Vivo Knee Loads” dataset. ResultsPerformance criteria based on either stress or muscle activation (either unscaled or scaled by muscle volume), both to a power of 3 or 4, resulted in the best agreement between measured and simulated values. These performance criteria outperformed all other criteria in terms of agreement between simulated muscle excitations and EMG, whereas good agreement between measured and predicted contact forces was also observed for minimization of contact forces and metabolic energy. SignificanceGiven the large differences in accuracy obtained with different performance criteria (e.g., root mean square errors of contact forces differed up to 0.45 body weight), the results of our study are important to improve the validity of in silico assessment of joint loading.

Accounting for SNR in an Algorithm Using Wavelet Transform to Remove ECG Interference from EMG Signals

When the electromyography (EMG) signal is acquired from muscles in the torso, the electrocardiography (ECG) signal coming from heart activity can interfere. As a result, the EMG signal can be contaminated during data collection. In this paper, a technique based on discrete stationary wavelet transform (DSWT) is proposed to remove ECG interference from the EMG signal while taking into account the signal-to-noise ratio (SNR). The contaminated EMG signal is decomposed using 5-level DSWT with the Symlet wavelet function. The coefficients for levels 4 and 5, which are contaminated by ECG, are set to zero when their absolute values are less than or equal to a threshold determined for each SNR level. A clean EMG signal can then be obtained by inverse DSWT mapping of the new thresholded coefficients. We evaluated the performance of the proposed algorithm using simulated EMG contaminated with both simulated and real ECG signals, at 9 SNR levels from [Formula: see text]20 to 20[Formula: see text]dB with 5[Formula: see text]dB increments. The performance based on mean absolute error, correlation coefficient and relative error shows that the DSWT method is better than a high-pass filter.

Diagnosis of Neuromuscular Disorders Using DT-CWT and Rotation Forest Ensemble Classifier

Electromyographic (EMG) signals are utilized to analyze the neuromuscular disorders. Machine learning algorithms have been employed as a decision support system to detect neuromuscular disorders. EMG signals contain noise from different sources, such as electrical and electronic instruments and movement artifacts. In this paper, the multiscale principal component analysis (MSPCA) has been used to remove the impulsive noise from the EMG signals. Then, the dual-tree complex wavelet transform (DT-CWT) is utilized for feature extraction, and the rotation forest ensemble classifier is employed for the recognition of EMG signals. In addition, the performance of several classifiers with rotation forest has been studied. An efficient combination of DT-CWT and rotation forest achieved good performance, using tenfold cross validation regarding the total classification accuracy. Results are promising and showed that the rotation forest achieved an accuracy of 99.7% with clinical EMG signals using support vector machine and 96.6% with simulated EMG signals using the artificial neural network (ANN).

Foot Pronation Contributes to Altered Lower Extremity Loading After Long Distance Running.

This study presents an investigation of the changes in foot posture, joint kinematics, joint moments and joint contact forces in the lower extremity following a 5 k treadmill run. A relationship between knee and ankle joint loading and foot posture index (FPI) is developed. Twenty recreational male heel-strike runners participated in this study. All participants had a history of running exercise and were free from lower extremity injuries and foot deformities. Foot posture was assessed from a six-item FPI to quantitatively classify high supination to high pronation foot poses. The FPI is scored using a combination of observations and foot palpations. The three-dimensional marker trajectories, ground reaction force and surface electromyography (EMG) were recorded at pre and post-gait sessions conducted over-ground and 5 k running was conducted on a treadmill. Joint kinematics, joint moments and joint contact forces were computed in OpenSim. Simulated EMG activations were compared against experimental EMG to validate the model. A paired sample t-test was conducted using a 1D statistical parametric mapping method computed temporally. Hip joint moments and contact forces increased during initial foot contact following 5 k running. Knee abduction moment and superior-inferior knee contact force increased, whereas the knee extension moment decreased. Ankle plantarflexion moment and ankle contact forces increased during stance. FPI was found to be moderately correlated with peak knee and ankle moments. Recreational male runners presented increased static foot pronation after 5 k treadmill running. These findings suggest that following mid distance running foot pronation may be an early indicator of increased lower limb joint loading. Furthermore, the FPI may be used to quantify the changes in knee and ankle joint moments.

Simulation of electromyographic recordings following transcranial magnetic stimulation.

Transcranial magnetic stimulation (TMS) is a technique that enables noninvasive manipulation of neural activity and holds promise in both clinical and basic research settings. The effect of TMS on the motor cortex is often measured by electromyography (EMG) recordings from a small hand muscle. However, the details of how TMS generates responses measured with EMG are not completely understood. We aim to develop a biophysically detailed computational model to study the potential mechanisms underlying the generation of EMG signals following TMS. Our model comprises a feed-forward network of cortical layer 2/3 cells, which drive morphologically detailed layer 5 corticomotoneuronal cells, which in turn project to a pool of motoneurons. EMG signals are modeled as the sum of motor unit action potentials. EMG recordings from the first dorsal interosseous muscle were performed in four subjects and compared with simulated EMG signals. Our model successfully reproduces several characteristics of the experimental data. The simulated EMG signals match experimental EMG recordings in shape and size, and change with stimulus intensity and contraction level as in experimental recordings. They exhibit cortical silent periods that are close to the biological values and reveal an interesting dependence on inhibitory synaptic transmission properties. Our model predicts several characteristics of the firing patterns of neurons along the entire pathway from cortical layer 2/3 cells down to spinal motoneurons and should be considered as a viable tool for explaining and analyzing EMG signals following TMS. NEW & NOTEWORTHY A biophysically detailed model of EMG signal generation following transcranial magnetic stimulation (TMS) is proposed. Simulated EMG signals match experimental EMG recordings in shape and amplitude. Motor-evoked potential and cortical silent period properties match experimental data. The model is a viable tool to analyze, explain, and predict EMG signals following TMS.

A simulation study on the relation between the motor unit depth and action potential from multi-channel surface electromyography recordings

To investigate the spatial information of individual motor unit (MUs) using multi-channel surface electromyography (EMG) decomposition. The K-means clustering convolution kernel compensation (KmCKC) approach was employed to detect the innervation pulse trains (IPTs) from the simulated surface EMG signals, and the motor unit action potentials (MUAPs) were evaluated using the spike-triggered average (STA) technique. The relationships between the features of MUAP and MU depth were determinated with a least square fitting method. The errors of peak-to-peak (PTP) amplitude of reconstructed MUAPs were less than 5.73%, even with 0 dB signal-to-noise (SNR). The fitting errors with nonlinear model were less than 5.55% for SNRs higher than 20 dB. The results show that it is possible to provide a useful method for estimating MU depth from surface EMG recordings. It is expected to extend the applicability of surface EMG technique to more challenging clinical applications.

Which data should be tracked in forward-dynamic optimisation to best predict muscle forces in a pathological co-contraction case?

The choice of the cost-function for predicting muscle forces during a movement remains a challenge, especially in patients with neuromuscular disorders. Forward dynamics-based optimisations mainly track joint kinematics or torques, combined with a least-excitation criterion. Tracking marker trajectories and/or electromyography (EMG) has rarely been proposed. Our objective was to determine the best tracking objective-function to accurately predict the upper-limb muscle forces. A musculoskeletal model was created and EMG was simulated to obtain a reference movement – a shoulder abduction. A Gaussian noise (mean = 0; standard deviation = 15%) was added to the simulated EMG. Another noise – corresponding to the actual soft tissue artefacts (STA) of experimental shoulder abduction movements – was added to the trajectories of the markers placed on the model. Muscle forces were estimated from these noisy data, using forward dynamics assisted by six non-linear least-squared objective-functions. These functions involved the tracking of marker trajectories, joint angles or torques, with and without EMG-tracking. All six approaches used the same musculoskeletal model and were solved using a direct multiple shooting algorithm. Finally, the predicted joint angles, muscle forces and activations were compared to the reference values, using root-mean-square errors (RMSe) and biases. The force RMSe of the approach tracking both marker trajectories and EMG (18.45 ± 12.60 N) was almost five times lower than the one of the approach tracking only joint angles (82.37 ± 66.26 N) or torques (85.10 ± 116.40 N). Therefore, using EMG as a complementary tracking-data in forward dynamics seems to be promising for the estimation of muscle forces.

An EMG-marker tracking optimisation method for estimating muscle forces

Existing algorithms for estimating muscle forces mainly use least-activation criteria, which do not necessarily lead to physiologically consistent results. Our objective was to assess an innovative forward dynamics-based optimisation, assisted by both electromyography (EMG) and marker tracking, for estimating the upper-limb muscle forces. A reference movement was generated, and EMG was simulated to reproduce the desired joint kinematics. Random noise was added to both simulated EMG and marker trajectories in order to create 30 trials. Then, muscle forces were estimated using (1) the innovative EMG-marker tracking forward optimisation, (2) a marker tracking forward optimisation with a least-excitation criterion, and (3) static optimisation with a least-activation criterion. Approaches (1) and (2) were solved using a direct multiple shooting algorithm. Finally, reference and estimated joint angles and muscle forces for the three optimisations were statistically compared using root-mean-square errors (RMSEs), biases, and statistical parametric mapping. The joint angles RMSEs were qualitatively similar across the three optimisations: (1) $1.63 \pm 0.51$ °; (2) $2.02 \pm 0.64$ °; (3) $0.79 \pm 0.38$ °. However, the muscle forces RMSE for the EMG-marker tracking optimisation ( $20.39 \pm 13.24$ N) was about seven times smaller than those resulting from the marker tracking ( $124.22 \pm 118.22$ N) and static ( $148.15 \pm 94.01$ N) optimisations. The originality of this novel approach is close tracking of both simulated EMG and marker trajectories in the same objective function, using forward dynamics. Therefore, the presented EMG-marker tracking optimisation led to accurate muscle forces estimations.

A Novel Unsupervised Adaptive Learning Method for Long-Term Electromyography (EMG) Pattern Recognition.

Performance degradation will be caused by a variety of interfering factors for pattern recognition-based myoelectric control methods in the long term. This paper proposes an adaptive learning method with low computational cost to mitigate the effect in unsupervised adaptive learning scenarios. We presents a particle adaptive classifier (PAC), by constructing a particle adaptive learning strategy and universal incremental least square support vector classifier (LS-SVC). We compared PAC performance with incremental support vector classifier (ISVC) and non-adapting SVC (NSVC) in a long-term pattern recognition task in both unsupervised and supervised adaptive learning scenarios. Retraining time cost and recognition accuracy were compared by validating the classification performance on both simulated and realistic long-term EMG data. The classification results of realistic long-term EMG data showed that the PAC significantly decreased the performance degradation in unsupervised adaptive learning scenarios compared with NSVC (9.03% ± 2.23%, p < 0.05) and ISVC (13.38% ± 2.62%, p = 0.001), and reduced the retraining time cost compared with ISVC (2 ms per updating cycle vs. 50 ms per updating cycle).

Analysis of statistical and standard algorithms for detecting muscle onset with surface electromyography.

The timing of muscle activity is a commonly applied analytic method to understand how the nervous system controls movement. This study systematically evaluates six classes of standard and statistical algorithms to determine muscle onset in both experimental surface electromyography (EMG) and simulated EMG with a known onset time. Eighteen participants had EMG collected from the biceps brachii and vastus lateralis while performing a biceps curl or knee extension, respectively. Three established methods and three statistical methods for EMG onset were evaluated. Linear envelope, Teager-Kaiser energy operator + linear envelope and sample entropy were the established methods evaluated while general time series mean/variance, sequential and batch processing of parametric and nonparametric tools, and Bayesian changepoint analysis were the statistical techniques used. Visual EMG onset (experimental data) and objective EMG onset (simulated data) were compared with algorithmic EMG onset via root mean square error and linear regression models for stepwise elimination of inferior algorithms. The top algorithms for both data types were analyzed for their mean agreement with the gold standard onset and evaluation of 95% confidence intervals. The top algorithms were all Bayesian changepoint analysis iterations where the parameter of the prior (p0) was zero. The best performing Bayesian algorithms were p0 = 0 and a posterior probability for onset determination at 60–90%. While existing algorithms performed reasonably, the Bayesian changepoint analysis methodology provides greater reliability and accuracy when determining the singular onset of EMG activity in a time series. Further research is needed to determine if this class of algorithms perform equally well when the time series has multiple bursts of muscle activity.

Extraction of EMG signal in a software compatible format from an online database using WFDB package

Summary The electromyography (EMG) is an electro diagnostic procedure for evaluating the electrical activity of skeletal muscles. It is very difficult to deduce the structure and activation of surrounding muscle tissue from needle-detected EMG signals and thus the extraction of information regarding presence, absence or severity of disease is not possible. The muscle morphology, activation and similarity with measured electromyographic signal features can be explored by comparison with simulated EMG signal. Evaluation of the data using such reference model will provide better understanding of the properties of muscles and their relationship to EMG signals. As the difficulty lies in the non-availability of standard EMG signal, the present work thus focuses on MATLAB implementation of needle detected EMG model introduced by Hamilton-Wright and Stashuk in 2005. The first step of analysis involves conversion of raw EMG signal from .dat format to .txt/.mat format which is achieved by using WFDB (Waveform Database) Software Package. In this research work standard simulated signals available in S002 dataset ( Hamilton-Wright and Stashuk, 2005 ) are extracted and the proposed procedure is presented in a detailed manner.

Elimination of Baseline Fluctuation in EMGSignal Using Digital Filter

Electromyography (EMG) is the study of the skeletal muscle function and the contraction of the skeletal muscle result in the generation of action potential in the individual muscle fibers, it from get information in the diagnosis of neuromuscular disorders. Suitable removal of baseline fluctuation of EMG signals is very important issue when recoding EMG signals as it may be degrade quality and quantitative analysis. In this paper focus on digital filter design technique for removal of noise present in the baseline of the EMG signal. The method is to estimate the spectral contents of the BLF, and then to use this estimation to design a high pass chebyshev filter by using Bilinear Transformation that cancel the BLF in the EMG signal. This technique on real and simulated EMG signal gives their advantages and disadvantages in term of both visual inspection and merit figures. We analyzed EMG signal from the muscles in a healthy subjects at low force level, using the concentric needle electrode.

Power spectrum of the rectified EMG: when and why is rectification beneficial for identifying neural connectivity?

Objective. The identification of common oscillatory inputs to motor neurons in the electromyographic (EMG) signal power spectrum is often preceded by EMG rectification for enhancing the low-frequency oscillatory components. However, rectification is a nonlinear operator and its influence on the EMG signal spectrum is not fully understood. In this study, we aim at determining when EMG rectification is beneficial in the study of oscillatory inputs to motor neurons. Approach. We provide a full mathematical description of the power spectrum of the rectified EMG signal and the influence of the average shape of the motor unit action potentials on it. We also provide a validation of these theoretical results with both simulated and experimental EMG signals. Main results. Simulations using an advanced computational model and experimental results demonstrated the accuracy of the theoretical derivations on the effect of rectification on the EMG spectrum. These derivations proved that rectification is beneficial when assessing the strength of low-frequency (delta and alpha bands) common synaptic inputs to the motor neurons, when the duration of the action potentials is short, and when the level of cancellation is relatively low. On the other hand, rectification may distort the estimation of common synaptic inputs when studying higher frequencies (beta and gamma), in a way dependent on the duration of the action potentials, and may introduce peaks in the coherence function that do not correspond to physiological shared inputs. Significance. This study clarifies the conditions when rectifying the surface EMG is appropriate for studying neural connectivity.

Detection of and Compensation for EMG Disturbances for Powered Lower Limb Prosthesis Control.

Myoelectric pattern recognition algorithms have been proposed for the control of powered lower limb prostheses, but electromyography (EMG) signal disturbances remain an obstacle to clinical implementation. To address this problem, we used a log-likelihood metric to detect simulated EMG disturbances and real disturbances acquired from EMG containing electrode shift. We found that features extracted from disturbed EMG have much lower log likelihoods than those from undisturbed signals and can be detected using a single threshold acquired from the training data. We designed a linear discriminant analysis (LDA) classifier that uses the log likelihood to decide between using a combination of EMG and mechanical sensors and using mechanical sensors only, to predict locomotion modes. When EMG contained disturbances, our classifier detected those disturbances and disregarded EMG data. Our classifier had significantly lower errors than a standard LDA classifier in the presence of EMG disturbances. The log-likelihood classifier had a low false positive threshold, and thus did not perform significantly differently from the standard LDA classifier when EMG did not contain disturbances. The log-likelihood threshold could also be applied to individual EMG channels, enabling specific channels containing EMG disturbances to be appropriately ignored when making locomotion mode predictions.