Phasic and tonic coupling between EEG and EMG demonstrated with independent component analysis

Network modeling and analysis of lumbar muscle surface EMG signals during flexion-extension in individuals with and without low back pain

In this paper, we propose modeling the activity coordination network between lumbar muscles using surface electromyography (sEMG) signals and performing the network analysis to compare the lumbar muscle coordination patterns between patients with low back pain (LBP) and healthy control subjects. Ten healthy subjects and eleven LBP patients were asked to perform flexion-extension task, and the sEMG signals were recorded. Both the subject-level and the group-level PC(fdr) algorithms are applied to learn the sEMG coordination networks with the error-rate being controlled. The network features are further characterized in terms of network symmetry, global efficiency, clustering coefficient and graph modules. The results indicate that the networks representing the normal group are much closer to the order networks and clearly exhibit globally symmetric patterns between the left and right sEMG channels. While the coordination activities between sEMG channels for the patient group are more likely to cluster locally and the group network shows the loss of global symmetric patterns. As a complementary tool to the physical and anatomical analysis, the proposed network analysis approach allows the visualization of the muscle coordination activities and the extraction of more informative features from the sEMG data for low back pain studies.

Identification of hand and finger movements using multi run ICA of surface electromyogram

Surface electromyogram (sEMG) based control of prosthesis and computer assisted devices can provide the user with near natural control. Unfortunately there is no suitable technique to classify sEMG when the there are multiple active muscles such as during finger and wrist flexion due to cross-talk. Independent Component Analysis (ICA) to decompose the signal into individual muscle activity has been demonstrated to be useful. However, ICA is an iterative technique that has inherent randomness during initialization. The average improvement in classification of sEMG that was separated using ICA was very small, from 60% to 65%. To overcome this problem associated with randomness of initialization, multi-run ICA (MICA) based sEMG classification system has been proposed and tested. MICA overcame the shortcoming and the results indicate that using MICA, the accuracy of identifying the finger and wrist actions using sEMG was 99%.

Validation of regression-based myogenic correction techniques for scalp and source-localized EEG

EEG and EEG source-estimation are susceptible to electromyographic artifacts (EMG) generated by the cranial muscles. EMG can mask genuine effects or masquerade as a legitimate effect-even in low frequencies, such as alpha (8-13 Hz). Although regression-based correction has been used previously, only cursory attempts at validation exist, and the utility for source-localized data is unknown. To address this, EEG was recorded from 17 participants while neurogenic and myogenic activity were factorially varied. We assessed the sensitivity and specificity of four regression-based techniques: between-subjects, between-subjects using difference-scores, within-subjects condition-wise, and within-subject epoch-wise on the scalp and in data modeled using the LORETA algorithm. Although within-subject epoch-wise showed superior performance on the scalp, no technique succeeded in the source-space. Aside from validating the novel epoch-wise methods on the scalp, we highlight methods requiring further development.

Hidden Markov multivariate autoregressive (HMM-mAR) modeling framework for surface electromyography (sEMG) data

Surface electromyographic (sEMG) analysis is complicated by the fact that the data are inherently non-stationary. To deal with this and to determine muscle activity patterns during reaching movements, we proposed modeling sEMG with a hidden Markov model-multivariate autoregressive (HMM-mAR) framework. The classification between healthy and stroke subjects was performed using structural features extracted from HMM-mAR models. Both the raw and carrier data produced excellent classification performance. The proposed method represents a fundamental departure from most existing methods where only the amplitude is analyzed or the mAR coefficients are directly used for classification. In contrast, our analysis shows that structural features of the multivariate sEMG carrier data or the residuals after model fitting can enhance the classification of reaching movements.

Study of stroke condition and hand dominance using a hidden Markov, multivariate autoregressive (HMM-mAR) network framework

To investigate the effects of stroke and hand dominance on muscle association patterns during reaching movements, we applied the hidden Markov model, multivariate autoregressive (HMM-mAR) framework to real sEMG recordings from healthy and stroke subjects performing reaching tasks. Statistical analysis is performed to construct subject- and group-level muscle connectivity networks. Associating structural features are extracted for subsequent classification of reaching movements. The HMM-mAR framework is shown to be able to consistently segments each reaching movement into the initial phase and the full-movement phase. The inferred muscle networks illustrate that healthy and stroke subjects use distinguishably different muscle synergies during the initial phase. The classification results further confirm that structural features extracted from the initial phase are useful in classifying subjects with differing stroke condition and handedness.

Improving functional magnetic resonance imaging motor studies through simultaneous electromyography recordings

Specially designed optoelectronic and data postprocessing methods are described that permit electromyography (EMG) of muscle activity simultaneous with functional MRI (fMRI). Hardware characterization and validation included simultaneous EMG and event-related fMRI in 17 healthy participants during either ankle (n = 12), index finger (n = 3), or wrist (n = 2) contractions cued by visual stimuli. Principal component analysis (PCA) and independent component analysis (ICA) were evaluated for their ability to remove residual fMRI gradient-induced signal contamination in EMG data. Contractions of ankle tibialis anterior and index finger abductor were clearly distinguishable, although observing contractions from the wrist flexors proved more challenging. To demonstrate the potential utility of simultaneous EMG and fMRI, data from the ankle experiments were analyzed using two approaches: 1) assuming contractions coincided precisely with visual cues, and 2) using EMG to time the onset and offset of muscle contraction precisely for each participant. Both methods produced complementary activation maps, although the EMG-guided approach recovered more active brain voxels and revealed activity better in the basal ganglia and cerebellum. Furthermore, numerical simulations confirmed that precise knowledge of behavioral responses, such as those provided by EMG, are much more important for event-related experimental designs compared to block designs. This simultaneous EMG and fMRI methodology has important applications where the amplitude or timing of motor output is impaired, such as after stroke.

Cortical muscle coupling in Parkinson's disease (PD) bradykinesia

OBJECTIVES

To determine if novel methods establishing patterns in EEG-EMG coupling can infer subcortical influences on the motor cortex, and the relationship between these subcortical rhythms and bradykinesia.

BACKGROUND

Previous work has suggested that bradykinesia may be a result of inappropriate oscillatory drive to the muscles. Typically, the signal processing method of coherence is used to infer coupling between a single channel of EEG and a single channel of rectified EMG, which demonstrates 2 peaks during sustained contraction: one, approximately 10 Hz, which is pathologically increased in PD, and a approximately 30 Hz peak which is decreased in PD, and influenced by pharmacological manipulation of GABAA receptors in normal subjects.

MATERIALS AND METHODS

We employed a novel multiperiodic squeezing paradigm which also required simultaneous movements. Seven PD subjects (on and off L-Dopa) and five normal subjects were recruited. Extent of bradykinesia was inferred by reduced relative performance of the higher frequencies of the squeezing paradigm and UPDRS scores. We employed Independent Component Analysis (ICA) and Empirical Mode Decomposition (EMD) to determine EEG/EMG coupling.

RESULTS

Corticomuscular coupling was detected during the continually changing force levels. Different components included those over the primary motor cortex (ipsilaterally and contralaterally) and over the midline. Subjects with greater bradykinesia had a tendency towards increased approximately 10 Hz coupling and reduced approximately 30 Hz coupling that was erratically reversed with L-dopa.

CONCLUSIONS

These results suggest that lower approximately 10 Hz peak may represent pathological oscillations within the basal ganglia which may be a contributing factor to bradykinesia in PD.

The Brain Computer Interface Using Flash Visual Evoked Potential and Independent Component Analysis

In this study flashing stimuli, such as digits or letters, are displayed on a LCD screen to induce flash visual evoked potentials (FVEPs). The aim of the proposed interface is to generate desired strings while one stares at target stimulus one after one. To effectively extract visually-induced neural activities with superior signal-to-noise ratio, independent component analysis (ICA) is employed to decompose the measured EEG and task-related components are subsequently selected for data reconstruction. In addition, all the flickering sequences are designed to be mutually independent in order to remove the contamination induced by surrounding non-target stimuli from the ICA-recovered signals. Since FVEPs are time-locked and phase-locked to flash onsets of gazed stimulus, segmented epochs from ICA-recovered signals based on flash onsets of gazed stimulus will be sharpen after averaging whereas those based on flash onsets of non-gazed stimuli will be suppressed after averaging. The stimulus inducing the largest averaged FVEPs is identified as the gazed target and corresponding digit or letter is sent out. Five subjects were asked to gaze at each stimulus. The mean detection accuracy resulted from averaging 15 epochs was 99.7%. Another experiment was to generate a specified string '0287513694E'. The mean accuracy and information transfer rates were 83% and 23.06 bits/min, respectively.

Multivariate AR modeling of electromyography for the classification of upper arm movements

OBJECTIVE

We compared the performance of two feature extraction methods for multichannel electromyography (EMG) based arm movement classification. One method was to use a scalar autoregressive model (sAR) for each channel. Another was to model all channels as a whole by a multivariate AR model (mAR).

METHODS

The classified arm movements included elbow flexion, elbow extension, forearm pronation and internal shoulder rotation. Six-channel bipolar EMG signals were collected from four electrodes fixed on the biceps, triceps, brachioradialis and deltoid. Fifteen two-channel and four three-channel configurations were formed out of these six-channel signals for a comparison of different channel combinations. Leave-one-out cross-validation was adopted for evaluating the classification performance using a parametric statistical classifier.

RESULTS

We processed a total of 216 EMG segments obtained from repeated 18 performances by three normal subjects. mAR model based feature set achieved a better classification accuracy than sAR did for each configuration. Moreover, significance of improvement was greater than 0.95 for those configurations which consisted of EMG channels that were close spatially.

CONCLUSIONS

The stronger the cross-correlation among EMG channels the more improvement of classification accuracy one would expect from using a mAR model.

Feature selection of stabilometric parameters based on principal component analysis

This study addresses the challenge of identifying the features of the Centre of pressure (COP) trajectory that are most sensitive to postural performance, with the aim of avoiding redundancy and allowing a straightforward interpretation of the results. Postural sway in 50 young, healthy subjects was measured by a force platform. Thirty-seven stabilometric parameters were computed from the one-dimensional and two-dimensional COP time series. After normalisation to the relevant biomechanical factors, by means of multiple regression models, a feature selection process was performed based on principal component analysis. Results suggest that COP two-dimensional time series can be primarily characterised by four parameters, describing the size of the COP path over the support surface; the principal sway direction; and the shape and bandwidth of the power spectral density plot. COP one-dimensional time series (antero-posterior (AP) and medio-lateral (ML)) can be characterised by six parameters describing COP dispersion along the AP direction; mean velocity along the ML and AP directions; the contrast between ML and AP regulatory activity; and two parameters describing the spectral characteristics of the COP along the AP direction. On the basis of the results obtained, some guidelines are suggested for the choice of stabilometric parameters to use, with the aim of promoting standardisation in quantitative posturography.

ICA-based spatiotemporal approach for single-trial analysis of postmovement MEG beta synchronization⋆

The extraction of event-related oscillatory neuromagnetic activities from single-trial measurement is challenging due to the non-phase-locked nature and variability from trial to trial. The present study presents a method based on independent component analysis (ICA) and the use of a template-based correlation approach to extract Rolandic beta rhythm from magnetoencephalographic (MEG) measurements of right finger lifting. A single trial recording was decomposed into a set of coupled temporal independent components and corresponding spatial maps using ICA and the reactive beta frequency band for each trial identified using a two-spectrum comparison between the postmovement interval and a reference period. Task-related components survived dual criteria of high correlation with both the temporal and the spatial templates with an acceptance rate of about 80%. Phase and amplitude information for noise-free MEG beta activities were preserved not only for optimal calculation of beta rebound (event-related synchronization) but also for profound penetration into subtle dynamics across trials. Given the high signal-to-noise ratio (SNR) of this method, various methods of source estimation were used on reconstructed single-trial data and the source loci coherently anchored in the vicinity of the primary motor area. This method promises the possibility of a window into the intricate brain dynamics of motor control mechanisms and the cortical pathophysiology of movement disorder on a trial-by-trial basis.

Non-invasive monitoring of functionally distinct muscle activations during swallowing

OBJECTIVES

Dysphagia is an important consequence of many diseases. As some of the muscles of deglutition tend to be deep to the surface, needle electrodes are typically used, but this limits the number of muscles that can be simultaneously recorded. Since control of swallowing involves central pattern generators (CPGs) which distribute commands to several muscles, monitoring several muscles simultaneously is desirable. Here we describe a novel method, based on computing the independent components (ICs) of the simultaneous sEMG recordings (Muscle Nerve Suppl 9 (2000) 9) to detect the underlying functional muscle activations during swallowing using only surface EMG (sEMG) electrodes.

METHODS

Seven normal subjects repeatedly swallowed liquids of varying consistency while sEMG was recorded from 15 electrodes from the face and throat. Active areas of EMG were excised from the recordings and the ICs of the sEMG were calculated.

RESULTS

The ICs demonstrated less swallow-to-swallow variability than the raw sEMG. The ICs, each consisting of a unique temporal waveform and a spatial distribution, provided a means to segregate the complex sequence of muscle activation into rigorously defined separate functional units. The temporal profiles of the ICs and their spatial distribution were consistent with prior needle EMG studies of the cricopharyngeal, superior pharyngeal constrictor, submental and possibly arytenoid muscles. Other components appeared to correspond to EKG artifact contaminating the EMG recordings, laryngeal excursion, tongue movement and activation of the buccal and/or masseter musculature At least two of the components were affected by the consistency of the liquids swallowed. Re-testing one subject a week later demonstrated good intertrial reliability.

CONCLUSIONS

We propose that the ICs of the sEMG provide a non-invasive means to assess the complex muscle sequence activation of deglutition.