Independence: a new criterion for the analysis of the electromagnetic fields in the global brain?

EEG Signal Complexity Measurements to Enhance BCI-Based Stroke Patients' Rehabilitation

The second leading cause of death and one of the most common causes of disability in the world is stroke. Researchers have found that brain-computer interface (BCI) techniques can result in better stroke patient rehabilitation. This study used the proposed motor imagery (MI) framework to analyze the electroencephalogram (EEG) dataset from eight subjects in order to enhance the MI-based BCI systems for stroke patients. The preprocessing portion of the framework comprises the use of conventional filters and the independent component analysis (ICA) denoising approach. Fractal dimension (FD) and Hurst exponent (Hur) were then calculated as complexity features, and Tsallis entropy (TsEn) and dispersion entropy (DispEn) were assessed as irregularity parameters. The MI-based BCI features were then statistically retrieved from each participant using two-way analysis of variance (ANOVA) to demonstrate the individuals' performances from four classes (left hand, right hand, foot, and tongue). The dimensionality reduction algorithm, Laplacian Eigenmap (LE), was used to enhance the MI-based BCI classification performance. Utilizing k-nearest neighbors (KNN), support vector machine (SVM), and random forest (RF) classifiers, the groups of post-stroke patients were ultimately determined. The findings show that LE with RF and KNN obtained 74.48% and 73.20% accuracy, respectively; therefore, the integrated set of the proposed features along with ICA denoising technique can exactly describe the proposed MI framework, which may be used to explore the four classes of MI-based BCI rehabilitation. This study will help clinicians, doctors, and technicians make a good rehabilitation program for people who have had a stroke.

EEG artifact removal using sub-space decomposition, nonlinear dynamics, stationary wavelet transform and machine learning algorithms

Blind source separation (BSS) methods have received a great deal of attention in electroencephalogram (EEG) artifact elimination as they are routine and standard signal processing tools to remove artifacts and reserve desired neural information. On the other hand, a classifier should follow BSS methods to automatically identify artifactual sources and remove them in the following steps. In addition, removing all detected artifactual components leads to loss of information since some desired information related to neural activity leaks to these sources. So, an approach should be employed to detect and suppress the artifacts and reserve neural activity. This study introduces a novel method based on EEG and Poincare planes in the phase space to detect artifactual components estimated by second-order blind identification (SOBI). Artifacts are detected using a mixture of well-known conventional classifiers and were removed employing stationary wavelet transform (SWT) to reserve neural information. The proposed method is a combination of signal processing techniques and machine learning algorithms, including multi-layer perceptron (MLP), K-nearest neighbor (KNN), naïve Bayes, and support vector machine (SVM) which have significant results while applying our proposed method to different scenarios. Simulated, semi-simulated, and real EEG signals are employed to evaluate the proposed method, and several evaluation criteria are calculated. We achieved acceptable results, for example, 98% average accuracy and 97% average sensitivity in artifactual EEG component detection or about 2% as mean square error in EEG reconstruction after artifact removal. Results showed that the proposed method is effective and can be used in future studies as we have considered different real-world scenarios to evaluate it.

Automatic Artifact Removal in EEG of Normal and Demented Individuals Using ICA-WT during Working Memory Tasks

Characterizing dementia is a global challenge in supporting personalized health care. The electroencephalogram (EEG) is a promising tool to support the diagnosis and evaluation of abnormalities in the human brain. The EEG sensors record the brain activity directly with excellent time resolution. In this study, EEG sensor with 19 electrodes were used to test the background activities of the brains of five vascular dementia (VaD), 15 stroke-related patients with mild cognitive impairment (MCI), and 15 healthy subjects during a working memory (WM) task. The objective of this study is twofold. First, it aims to enhance the recorded EEG signals using a novel technique that combines automatic independent component analysis (AICA) and wavelet transform (WT), that is, the AICA–WT technique; second, it aims to extract and investigate the spectral features that characterize the post-stroke dementia patients compared to the control subjects. The proposed AICA–WT technique is a four-stage approach. In the first stage, the independent components (ICs) were estimated. In the second stage, three-step artifact identification metrics were applied to detect the artifactual components. The components identified as artifacts were marked as critical and denoised through DWT in the third stage. In the fourth stage, the corrected ICs were reconstructed to obtain artifact-free EEG signals. The performance of the proposed AICA–WT technique was compared with those of two other techniques based on AICA and WT denoising methods using cross-correlation XCorr and peak signal to noise ratio (PSNR) (ANOVA, p ˂ 0.05). The AICA–WT technique exhibited the best artifact removal performance. The assumption that there would be a deceleration of EEG dominant frequencies in VaD and MCI patients compared with control subjects was assessed with AICA–WT (ANOVA, p ˂ 0.05). Therefore, this study may provide information on post-stroke dementia particularly VaD and stroke-related MCI patients through spectral analysis of EEG background activities that can help to provide useful diagnostic indexes by using EEG signal processing.

Advanced Electroencephalogram Processing

The study of electroencephalography (EEG) data can involve independent component analysis and further clustering of the components according to relation of the components to certain processes in a brain or to external sources of electricity such as muscular motion impulses, electrical fields inducted by power mains, electrostatic discharges, etc. At present, known methods for clustering of components are costly because require additional measurements with magnetic-resonance imaging (MRI), for example, or have accuracy restrictions if only EEG data is analyzed. A new method and algorithm for automatic clustering of physiologically similar but statistically independent EEG components is described in this paper. Developed clustering algorithm has been compared with algorithms implemented in the EEGLab toolbox. The paper contains results of algorithms testing on real EEG data obtained under two experimental tasks: voluntary movement control under conditions of stop-signal paradigm and syntactical error recognition in written sentences. The experimental evaluation demonstrated more than 90% correspondence between the results of automatic clustering and clustering made by an expert physiologist.

MEG source imaging method using fast L1 minimum-norm and its applications to signals with brain noise and human resting-state source amplitude images

The present study developed a fast MEG source imaging technique based on Fast Vector-based Spatio-Temporal Analysis using a L1-minimum-norm (Fast-VESTAL) and then used the method to obtain the source amplitude images of resting-state magnetoencephalography (MEG) signals for different frequency bands. The Fast-VESTAL technique consists of two steps. First, L1-minimum-norm MEG source images were obtained for the dominant spatial modes of sensor-waveform covariance matrix. Next, accurate source time-courses with millisecond temporal resolution were obtained using an inverse operator constructed from the spatial source images of Step 1. Using simulations, Fast-VESTAL's performance was assessed for its 1) ability to localize multiple correlated sources; 2) ability to faithfully recover source time-courses; 3) robustness to different SNR conditions including SNR with negative dB levels; 4) capability to handle correlated brain noise; and 5) statistical maps of MEG source images. An objective pre-whitening method was also developed and integrated with Fast-VESTAL to remove correlated brain noise. Fast-VESTAL's performance was then examined in the analysis of human median-nerve MEG responses. The results demonstrated that this method easily distinguished sources in the entire somatosensory network. Next, Fast-VESTAL was applied to obtain the first whole-head MEG source-amplitude images from resting-state signals in 41 healthy control subjects, for all standard frequency bands. Comparisons between resting-state MEG sources images and known neurophysiology were provided. Additionally, in simulations and cases with MEG human responses, the results obtained from using conventional beamformer technique were compared with those from Fast-VESTAL, which highlighted the beamformer's problems of signal leaking and distorted source time-courses.

Dynamic changes of ICA-derived EEG functional connectivity in the resting state

An emerging issue in neuroscience is how to identify baseline state(s) and accompanying networks termed "resting state networks" (RSNs). Although independent component analysis (ICA) in fMRI studies has elucidated synchronous spatiotemporal patterns during cognitive tasks, less is known about the changes in EEG functional connectivity between eyes closed (EC) and eyes open (EO) states, two traditionally used baseline indices. Here we investigated healthy subjects (n = 27) in EC and EO employing a four-step analytic approach to the EEG: (1) group ICA to extract independent components (ICs), (2) standardized low-resolution tomography analysis (sLORETA) for cortical source localization of IC network nodes, followed by (3) graph theory for functional connectivity estimation of epochwise IC band-power, and (4) circumscribing IC similarity measures via hierarchical cluster analysis and multidimensional scaling (MDS). Our proof-of-concept results on alpha-band power demonstrate five statistically clustered groups with frontal, central, parietal, occipitotemporal, and occipital sources. Importantly, during EO compared with EC, graph analyses revealed two salient functional networks with frontoparietal connectivity: a more medial network with nodes in the mPFC/precuneus which overlaps with the "default-mode network" (DMN), and a more lateralized network comprising the middle frontal gyrus and inferior parietal lobule, coinciding with the "dorsal attention network" (DAN). Furthermore, a separate MDS analysis of ICs supported the emergence of a pattern of increased proximity (shared information) between frontal and parietal clusters specifically for the EO state. We propose that the disclosed component groups and their source-derived EEG functional connectivity maps may be a valuable method for elucidating direct neuronal (electrophysiological) RSNs in healthy people and those suffering from brain disorders.

Blind source separation to enhance spectral and non-linear features of magnetoencephalogram recordings. Application to Alzheimer's disease

This work studied whether a blind source separation (BSS) and component selection procedure could increase the differences between Alzheimer's disease (AD) patients and control subjects' spectral and non-linear features of magnetoencephalogram (MEG) recordings. MEGs were acquired with a 148-channel whole-head magnetometer from 62 subjects (36 AD patients and 26 controls), who were divided randomly into training and test sets. MEGs were decomposed using the algorithm for multiple unknown signals extraction (AMUSE). The extracted AMUSE components were characterised with two spectral--median frequency and spectral entropy (SpecEn)--and two non-linear features: Lempel-Ziv complexity (LZC) and sample entropy (SampEn). One-way analysis of variance with age as a covariate was applied to the training set to decide which components had the most significant differences between groups. Then, partial reconstructions of the MEGs were computed with these significant components. In the test set, the accuracy and area under the ROC curve (AUC) associated with each partial reconstruction of the MEGs were compared with the case where no BSS-preprocessing was applied. This preprocessing increased the AUCs between 0.013 and 0.227, while the accuracy for SpecEn, LZC and SampEn rose between 6.4% and 22.6%, improving the separation between AD patients and control subjects.

Artifact removal in magnetoencephalogram background activity with independent component analysis

The aim of this study was to assess whether independent component analysis (ICA) could be valuable to remove power line noise, cardiac, and ocular artifacts from magnetoencephalogram (MEG) background activity. The MEGs were recorded from 11 subjects with a 148-channel whole-head magnetometer. We used a statistical criterion to estimate the number of independent components. Then, a robust ICA algorithm decomposed the MEG epochs and several methods were applied to detect those artifacts. The whole process had been previously tested on synthetic data. We found that the line noise components could be easily detected by their frequency spectrum. In addition, the ocular artifacts could be identified by their frequency characteristics and scalp topography. Moreover, the cardiac artifact was better recognized by its skewness value than by its kurtosis one. Finally, the MEG signals were compared before and after artifact rejection to evaluate our method.

Comparison between ICA and wavelet-based denoising of single-trial evoked potentials

Analysis of evoked potentials (EPs) on a single-trial basis allows the study of the dynamical characteristics of brain activity. However, single-trial responses are buried into the more prominent ongoing electroencephalographic (EEG) activity, and thus advanced procedures are needed to obtain the activity only of the cortical generators that are activated by the experimental task under study. We compare the effectiveness of two methods at removing extraneous activity from single-trial EPs, namely, a recently-proposed iterative procedure based on independent component analysis (ICA) and wavelet denoising, using simulated data and actual EP recordings from normal subjects, more specifically the auditory N100-P200 complex.

Reliable detection of bilateral activation in human primary somatosensory cortex by unilateral median nerve stimulation

In non-human primates, a bilateral representation of unilaterally presented somatosensory information can be found at the lowest level of cortical processing as indicated by the presence of neurons with bilateral receptive fields in the hand region of primary somatosensory (SI) cortex. In humans, such bilateral activation of SI is considered controversial due to highly variable detection rates for the much weaker ipsilateral response across different studies (ranging from 3% to 100%). Second-order blind identification (SOBI) is a blind source separation algorithm that has been successfully used to isolate neuronal signals from functionally distinct brain regions, including the left- and right-SI. SOBI-aided extraction of left- and right-SI responses to median nerve stimulation from high-density EEG has been previously validated against the fMRI and MEG literature. Here, we applied SOBI to EEG data and examined whether relatively weaker ipsilateral activations could be reliably detected across subjects. In single subject analysis, statistically significant somatosensory evoked potentials (SEPs) in response to unilateral stimulation were detected from both SI contralateral to and SI ipsilateral to the side of stimulation. Furthermore, these ipsilateral responses were observed in both the left and right hemispheres of all 10 subjects studied. Together these results demonstrate that unilateral stimulation of the median nerve, whether applied to the left or right wrist, can activate both the left- and right-SI, raising the possibility that in humans, unilateral sensory input may be bilaterally represented at the lowest level of cortical processing.

Vector-based spatial–temporal minimum L1-norm solution for MEG

Minimum L1-norm solutions have been used by many investigators to analyze MEG responses because they provide high spatial resolution images. However, conventional minimum L1-norm approaches suffer from instability in spatial construction, and poor smoothness of the reconstructed source time-courses. Activity commonly "jumps" from one grid point to (usually) the neighboring grid points. Equivalently, the time-course of one specific grid point can show substantial "spiky-looking" discontinuity. In the present study, we present a new vector-based spatial-temporal analysis using a L1-minimum-norm (VESTAL). This approach is based on a principle of MEG physics: the magnetic waveforms in sensor-space are linear functions of the source time-courses in the imaging-space. Our computer simulations showed that VESTAL provides good reconstruction of the source amplitude and orientation, with high stability and resolution in both the spatial and temporal domains. "Spiky-looking" discontinuity was not observed in the source time-courses. Importantly, the simulations also showed that VESTAL can resolve sources that are 100% correlated. We then examined the performance of VESTAL in the analysis of human median-nerve MEG responses. The results demonstrated that this method easily distinguishes sources very spatially close to each other, including individual primary somatosensory areas (BA 1, 2, 3b), primary motor area (BA 4), and other regions in the somatosensory system (e.g., BA 5, 7, SII, SMA, and temporal-parietal junction) with high temporal stability and resolution. VESTAL's potential for obtaining information on source extent was also examined.

Independent Component Analysis Separates Spikes of Different Origin in the EEG

Independent component analysis (ICA) is a novel system that finds independent sources in recorded signals. Its usefulness in separating epileptiform activity of different origin has not been determined. The goal of this study was to demonstrate that ICA is useful for separating different spikes using samples of EEG of patients with focal epilepsy. Digital EEG samples from four patients with focal epilepsy were included. The patients had temporal (n = 2), centrotemporal (n = 1) or frontal spikes (n = 1). Twenty-six samples with two (or more) spikes from two different patients were created. The selection of the two spikes for each mixed EEG was performed randomly, trying to have all the different combinations and rejecting the mixture of two spikes from the same patient. Two different examiners studied the EEGs using ICA with JADE paradigm in Matlab platform, trying to separate and to identify the spikes. They agreed in the correct separation of the spikes in 24 of the 26 samples, classifying the spikes as frontal, temporal or centrotemporal, left or right sided. The demonstration of the possibility of detecting different artificially mixed spikes confirms that ICA may be useful in separating spikes or other elements in real EEGs.

Contrasting single-trial ERPs between experimental manipulations: Improving differentiability by blind source separation

Contrasting event-related potentials (ERPs) generated under different experimental conditions and inferring differential brain responses is widely practiced in cognitive neuroscience research. Traditionally, these contrasts and subsequent inferences have proceeded directly from ERPs measured at the scalp. For certain tasks, it is not unusual that ERPs from a subset of channels are given particular emphasis in data analysis, such as the channels displaying the maximum peak amplitude in regions of interest ("best sensors") or channels showing the largest averaged ERP waveform differences. With the aid of a blind source separation (BSS) algorithm, second-order blind identification (SOBI), which has been recently validated for its ability to recover correlated neuronal sources, we show that single-trial ERPs from previously validated neuronal sources were more distinguishable among different experimental manipulations than the single-trial ERPs measured at the comparable "best sensors". This suggests that by using validated SOBI-recovered neuronal sources, ERP researchers can improve the ability to detect differences in neuronal responses induced by experimental manipulations. Critically, our observations were made at the level of single trials, as opposed to the averaged ERP. Therefore, our conclusions are particularly relevant to phenomena involving trial-to-trial changes in brain activation, for example, rapid induction of brain plasticity and perceptual learning, as well as to the development of brain-computer interfaces. Similar advantages would also apply to analogous situations with magnetoencephalography (MEG).

Rejection of artifact sources in magnetoencephalogram background activity using independent component analysis

The aim of this pilot study was to assess the usefulness of independent component analysis (ICA) to detect cardiac artifacts and power line interferences in magnetoencephalogram (MEG) recordings. We recorded MEG signals from six subjects with a 148-channel whole-head magnetometer (MAGNES 2500 WH, 4D Neuroimaging). Epochs of 50 s with power line noise, cardiac, and ocular artifacts were selected for analysis. We applied a statistical criterion to determine the number of sources, and a robust ICA algorithm to decompose the MEG epochs. Skewness, kurtosis, and a spectral metric were used to mark the studied artifacts. We found that the power line interference could be easily detected by its frequency characteristics. Moreover, skewness outperformed kurtosis when identifying the cardiac artifact.

Recovery of correlated neuronal sources from EEG: the good and bad ways of using SOBI

Second-order blind identification (SOBI) is a blind source separation (BSS) algorithm that has been applied to MEG and EEG data collected during a range of sensory, motor, and cognitive tasks. SOBI can decompose mixtures of electric or magnetic signals by utilizing detailed temporal structures present in the continuously recorded signals. Successful decomposition critically depends on the choice of temporal delay parameters used for computing multiple covariance matrices. Here, we present empirical findings from high-density EEG data (128 channels) to show that SOBI's ability to recover correlated neuronal sources critically depends on the appropriate use of these temporal delay parameters. Specifically, we applied SOBI to EEG data collected during correlated activation of the left and right primary somatosensory cortices (SI). We show that separation of signals originating from the left and right SI is better achieved by using a large number and a wide range of temporal delays between a few and several hundred milliseconds when compared to results using various subsets of these delays. The paper also offers non-mathematician/engineer users a gentle introduction to the inner workings of SOBI.

Applying spatial distribution analysis techniques to classification of 3D medical images

OBJECTIVE

The objective of this paper is to classify 3D medical images by analyzing spatial distributions to model and characterize the arrangement of the regions of interest (ROIs) in 3D space.

METHODS AND MATERIAL

Two methods are proposed for facilitating such classification. The first method uses measures of similarity, such as the Mahalanobis distance and the Kullback-Leibler (KL) divergence, to compute the difference between spatial probability distributions of ROIs in an image of a new subject and each of the considered classes represented by historical data (e.g., normal versus disease class). A new subject is predicted to belong to the class corresponding to the most similar dataset. The second method employs the maximum likelihood (ML) principle to predict the class that most likely produced the dataset of the new subject.

RESULTS

The proposed methods have been experimentally evaluated on three datasets: synthetic data (mixtures of Gaussian distributions), realistic lesion-deficit data (generated by a simulator conforming to a clinical study), and functional MRI activation data obtained from a study designed to explore neuroanatomical correlates of semantic processing in Alzheimer's disease (AD).

CONCLUSION

Performed experiments demonstrated that the approaches based on the KL divergence and the ML method provide superior accuracy compared to the Mahalanobis distance. The later technique could still be a method of choice when the distributions differ significantly, since it is faster and less complex. The obtained classification accuracy with errors smaller than 1% supports that useful diagnosis assistance could be achieved assuming sufficiently informative historic data and sufficient information on the new subject.

Validation of SOBI components from high-density EEG

Second-order blind identification (SOBI) is a blind source separation (BSS) algorithm that can be used to decompose mixtures of signals into a set of components or putative recovered sources. Previously, SOBI, as well as other BSS algorithms, has been applied to magnetoencephalography (MEG) and electroencephalography (EEG) data. These BSS algorithms have been shown to recover components that appear to be physiologically and neuroanatomically interpretable. While some proponents of these algorithms suggest that fundamental discoveries about the human brain might be made through the application of these techniques, validation of BSS components has not yet received sufficient attention. Here we present two experiments for validating SOBI-recovered components. The first takes advantage of the fact that noise sources associated with individual sensors can be objectively validated independently from the SOBI process. The second utilizes the fact that the time course and location of primary somatosensory (SI) cortex activation by median nerve stimulation have been extensively characterized using converging imaging methods. In this paper, using both known noise sources and highly constrained and well-characterized neuronal sources, we provide validation for SOBI decomposition of high-density EEG data. We show that SOBI is able to (1) recover known noise sources that were either spontaneously occurring or artificially induced; (2) recover neuronal sources activated by median nerve stimulation that were spatially and temporally consistent with estimates obtained from previous EEG, MEG, and fMRI studies; (3) improve the signal-to-noise ratio (SNR) of somatosensory-evoked potentials (SEPs); and (4) reduce the level of subjectivity involved in the source localization process.

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.

On the choice of a sparse prior

An emerging paradigm analyses in what respect the properties of the nervous system reflect properties of natural scenes. It is hypothesized that neurons form sparse representations of natural stimuli: each neuron should respond strongly to some stimuli while being inactive upon presentation of most others. For a given network, sparse representations need fewest spikes, and thus the nervous system can consume the least energy. To obtain optimally sparse responses the receptive fields of simulated neurons are optimized. Algorithmically this is identical to searching for basis functions that allow coding for the stimuli with sparse coefficients. The problem is identical to maximizing the log likelihood of a generative model with prior knowledge of natural images. It is found that the resulting simulated neurons share most properties of simple cells found in primary visual cortex. Thus, forming optimally sparse representations is a very compact approach to describing simple cell properties. Many ways of defining sparse responses exist and it is widely believed that the particular choice of the sparse prior of the generative model does not significantly influence the estimated basis functions. Here we examine this assumption more closely. We include the constraint of unit variance of neuronal activity, used in most studies, into the objective functions. We then analyze learning on a database of natural (cat-cam) visual stimuli. We show that the effective objective functions are largely dominated by the constraint, and are therefore very similar. The resulting receptive fields show some similarities but also qualitative differences. Even for coefficient values for which the objective functions are dissimilar, the distributions of coefficients are similar and do not match the priors of the assumed generative model. In conclusion, the specific choice of the sparse prior is relevant, as is the choice of additional constraints, such as normalization of variance.