No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Evaluation of Artifact Subspace Reconstruction for Automatic Artifact Components Removal in Multi-Channel EEG Recordings
Abstract
Objective:
Artifact subspace reconstruction (ASR) is an automatic, online-capable, component-based method that can effectively remove transient or large-amplitude artifacts contaminating electroencephalographic (EEG) data. However, the effectiveness of ASR and the optimal choice of its parameter have not been systematically evaluated and reported, especially on actual EEG data.
Methods:
This study systematically evaluates ASR on twenty EEG recordings taken during simulated driving experiments. Independent component analysis (ICA) and an independent component classifier are applied to separate artifacts from brain signals to quantitatively assess the effectiveness of the ASR.
Results:
ASR removes more eye and muscle components than brain components. Moreover, even though some eye and muscle components retain after ASR cleaning, the power of their activities are reduced. Study results also showed that ASR cleaning improved the quality of a subsequent ICA decomposition.
Conclusions:
Empirical results show that the optimal ASR parameter is between 20 and 30, balancing between removing non-brain signals and retaining brain activities.
Significance:
With an appropriate choice of parameter, ASR can be a powerful and automatic artifact removal approach for offline data analysis or online real-time EEG applications such as clinical monitoring and brain-computer interfaces.
... A popular brain wave pattern that is used in EEG-based BMIs is event-related desynchronization (ERD) 4-6 . These ERDs manifest as an attenuation in alpha- (8)(9)(10)(11)(12)(13) or beta- (20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) band power when the user is engaged in limb movement. The suppression is typically found in sensors that are close to the somatotopic areas associated with the engaged limb. ...
... To remove the drift in the signals, they were high-pass filtered with a 4 th order zero phase Butterworth filter at a cutoff of 0.1 Hz. To remove brief deflections in the signals likely caused by head or sensor movements, artifact subspace reconstruction (ASR) was performed 27,28 . Time windows of 0.5 seconds that had a variance beyond 60 standard deviations were corrected with ASR. ...
... The Hamming window was used to reduce spectral leakage. From the power spectral density (PSD), spectral band power was extracted in the alpha band (8)(9)(10)(11)(12)(13) and in the beta band (20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30). ...
Brain Machine Interfaces (BMIs) are systems that can help restore motor function to individuals with paralysis. These systems utilize machine learning algorithms to detect how the user wants to control an assistive device from their brain activity. These algorithms require a training dataset to infer patterns between a user's brain waves and the control signal for the assistive device. These training datasets consist of brain activity recordings while participants perform a very repetitive motor task. We hypothesize that these training datasets can be compromised due to the high number of trials. We anticipate that the movement-related brain waves of interest attenuate over time due to neural efficiency, where the brain becomes more efficient with practice in a motor task. To explore this hypothesis, we used an open-access EEG dataset where participants performed a simple reach-and-lift task. From each trial, time windows associated with resting and movement periods were segmented. Alpha and beta-band spectral power was estimated for each of these epochs and event-related desynchronization (ERD) was estimated as the suppression in spectral power from rest to movement. These ERDs were compared between early and late trials in the dataset. We also used linear discriminant analysis to assess a machine learning algorithm's accuracy in classifying whether the time windows belonged to rest or movement based on the spectral power band. We found that earlier trials had stronger ERDs and larger classification accuracies, suggesting that the repetition in the motor task causes attenuation of movement-related brain wave patterns and reduces efficacy of the dataset. These results call for a reevaluation of BMI performance in datasets that have numerous trials and an exploration of strategies that can compensate for longitudinal changes in movement-related brain activity used for BMIs.
... Input α rhythm MAE in PSD ( σ) Wavelet Sym3 [43] Single channel 0.539 0.326 DWT-ANC [24] Single channel 0.416 0.251 EEMD-PCA [27] Single channel 0.669 0.504 EEMD-KPCA (Polynomial kernel of 3rd order) [44] Single channel 0.583 0.413 SSA-ANC [25] Single channel 0.167 0.053 DWT-ICA [23] Multichannel 0.031 0.011 FBSEEWT-LPATV [29] Single channel 0.029 0.020 EWT Single channel 9. The resemblance of the reconstructed EEG channel with the raw EEG channel is further evaluated in terms of the correlation coefficient index (CCI) measure; the channel by channel results have been compared with four existing state-of-the-art methods [15], [16], [20], [30], enumerated in Table VI. The Higher the CCI is, the better the method is at preserving the EEG information. ...
... The wavelet-ICA method [15] achieved average CCI of 0.77 for real 14-channel EEG signals. Likewise, the multichannel EEG eye-blink artifact suppression method, namely, the ASR method [20], achieves average CCI value of 0.85 when studied in the real EEG signals considered in this work. The state-of-the-art ICLabel method [16] removed the EOG artifacts from real EEG data, resulting in an average CCI value of 0.72. ...
... The proposed method takes 960 milliseconds for processing 64-channel EEG of duration 125 s. As can be seen in Fig. 18, the proposed method executes much faster than the traditional ASR method [20] for several real-world EEG sessions. On the other hand, the wavelet ICA method [15] has a mean execution time of 0.060 s for processing 12 channel EEG signal of 105 s duration. ...
The electroencephalogram (EEG) signals record electrical activities generated by the brain cells and are used as a state-of-the-art diagnosis tool for various neural disorders. However, the unwanted artifacts often contaminate the recorded EEG signals and disturb the interpretation of the neuronal activity. This paper aims to propose an efficient automatic method to eliminate the ocular artifacts (OAs) from the multi-channel EEG signals with novel frequency-spatial filtering. The method combines dictionary-based spatial filtering and frequency based signal decomposition method, namely empirical wavelet transform (EWT). The artifact dictionary needed for spatial filtering is isolated from the raw data by (1) selecting the contaminated channels and (2) frequency-domain filtering. More precisely, the δ-rhythms of identified highly contaminated channels are selected and placed into an artifact dictionary. Afterward, the δ-rhythms of multi-channel EEG signals are spatially filtered using the built dictionary to seclude the OAs within a limited number of components. Further, the artifact components are eliminated and clean δ-rhythms are recovered using inverse spatial filtering technique. Finally, the clean δ-rhythms are combined with other EEG rhythms to reconstruct the OA-free signals. The proposed method is applied to OA contaminated synthetic and real multi-channel EEG signals with a convincing performance as compared to state of the art approaches. The proposed method removes the OAs without affecting the background EEG information. The proposed method can ease sensor signal interpretation and further processing, e.g. for BCI applications.
... Here, we used a correlation threshold of 0.8. Moreover, we repaired nonstationary high-amplitude artifacts with the Artifact Subspace Reconstruction (ASR) algorithm [63], using optimal parameters as recommended in [64]. Specifically, the ASR, first, searches for the cleanest part of the data to use it as calibration data. ...
... The component-specific thresholds are based on the variance of the calibration data's PCs, multiplied by a user-specified factor which determines how many times the variance of the PCs of the windowed data should exceed the variance of the calibration data's PCs in order to be considered an artifact. Higher values of such factor imply a very conservative filtering, whereas lower values correspond to a very sensitive one, with the consequence of removing not only artifacts but also relevant information from the data [64]. In this work, we adopted a conservative factor of 30, which has been shown to remove the majority of the artifacts, while preserving the information contained in the data [64]. ...
... Higher values of such factor imply a very conservative filtering, whereas lower values correspond to a very sensitive one, with the consequence of removing not only artifacts but also relevant information from the data [64]. In this work, we adopted a conservative factor of 30, which has been shown to remove the majority of the artifacts, while preserving the information contained in the data [64]. We then visually inspected the EEG signal, to remove bad data not successfully repaired by the ASR. ...
The cortical network including the piriform (PC), orbitofrontal (OFC), and entorhinal (EC) cortices allows the complex processing of behavioral, cognitive, and context-related odor information and represents an access gate to the subcortical limbic regions. Among the several factors that influence odor processing, their hedonic content and gender differences play a relevant role. Here, we investigated how these factors influence EEG effective connectivity among the mentioned brain regions during emotional olfactory stimuli. To this aim, we acquired EEG data from twenty-one healthy volunteers, during a passive odor task of odorants with different valence. We used Dynamic Causal Modeling (DCM) for EEG and Parametric Empirical Bayes (PEB) to investigate the modulatory effects of odors’ valence on the connectivity strengths of the PC-EC-OFC network. Moreover, we controlled for the influence of arousal and gender on such modulatory effects. Our results highlighted the relevant role of the forward and backward PC-EC connections in odor’s brain processing. On the one hand, the EC-to-PC connection was inhibited by both pleasant and unpleasant odors, but not by the neutral one. On the other hand, the PC-to-EC forward connection was found to be modulated (posterior probability (Pp)>0.95) by the arousal level associated with an unpleasant odor. Finally, the whole network dynamics showed several significant gender-related differences (Pp>0.95) suggesting a better ability in odor discrimination for the female gender.
... computational burden [22]. Indeed, there is still no agreement on an optimal removal technique for all types of artifacts. ...
... The ASR process consists of three major steps [22]: 1) Extraction of reference data. A portion of the signal without artifacts is identified by calculating the rootmean-square (RMS) values on 1-second sliding windows for each channel. ...
... However, as reported in Section I, the majority of ASR-based works use standard parameters [24], [30]. Studies focusing on the optimal value of k, in particular, revealed that this value could be between 20 and 30, which is small enough to remove artifact and preserve most of brain information [27], [22]. ...
Electroencephalogram (EEG) plays a significant role in the analysis of cerebral activity, although the recorded electrical brain signals are always contaminated with artifacts. This represents the major issue limiting the use of EEG in daily life applications, as artifact removal process still remains a challenging task. Among the available methodologies, Artifact Subspace Reconstruction (ASR) is a promising tool that can effectively remove transient or large-amplitude artifacts. However, the effectiveness of ASR and the optimal choice of its parameters have been validated only for high-density EEG acquisitions. In this regard, the present study proposes an enhanced procedure for the optimal individuation of ASR parameters, in order to successfully remove artifact in low-density EEG acquisitions (down to four channels). The proposed method starts from the analysis of real EEG data, to generate a large semi-simulated dataset with similar characteristics. Through a fine-tuning procedure on this semi-simulated data, the proposed method identifies the optimal parameters to be used for artifact removal on real data. The results show that the algorithm achieves an efficient removal of artifacts preserving brain signal information, also in low-density EEG signals, thus favoring the adoption of EEG also for more portable and/or daily-life applications.
... The narrowband EEG data were then passed through multiple data cleaning stages to remove the artifacts from the data. Artifact Subspace Reconstruction (ASR) is a widely used method for cleaning EEG data and uses the variance-based algorithm presented in [40]. This algorithm creates a statistical model of the clean EEG portion in the data and applies principal component analysis (PCA) to new incoming raw signal and transforms it into the principal component (PC) space. ...
... First, continuous EEG data were filtered via a FIR filter between 0.5 and 40 Hz using a Hamming window with the help of the pop_eegfiltnew function in EEGLAB. After that, ASR was used to reconstruct the artifact portion of the data with clean data [40]. This method removes the non-stationary high-variance signal and, then, reconstructs the missing data using a spatial mixing matrix. ...
Neurofeedback, an operant conditioning neuromodulation technique, uses information from brain activities in real-time via brain–computer interface (BCI) technology. This technique has been utilized to enhance the cognitive abilities, including working memory performance, of human beings. The aims of this study are to investigate how alpha neurofeedback can improve working memory performance in healthy participants and to explore the underlying neural mechanisms in a working memory task before and after neurofeedback. Thirty-six participants divided into the NFT group and the control group participated in this study. This study was not blinded, and both the participants and the researcher were aware of their group assignments. Increasing power in the alpha EEG band was used as a neurofeedback in the eyes-open condition only in the NFT group. The data were collected before and after neurofeedback while they were performing the N-back memory task (N = 1 and N = 2). Both groups showed improvement in their working memory performance. There was an enhancement in the power of their frontal alpha and beta activities with increased working memory load (i.e., 2-back). The experimental group showed improvements in their functional connections between different brain regions at the theta level. This effect was absent in the control group. Furthermore, brain hemispheric lateralization was found during the N-back task, and there were more intra-hemisphere connections than inter-hemisphere connections of the brain. These results suggest that healthy participants can benefit from neurofeedback and from having their brain networks changed after the training.
... Increasing the cut-off parameter increases the threshold values and leads to very conservative filtering. On the other hand, decreasing the cut-off parameter decreases the threshold values and leads to very sensitive filtering, causing not only the rejection of artifacts but also the loss of relevant information from the data 40 . Here, we adopted a conservative cut-off parameter of 30, which allows for the rejection of the majority of artifacts while preserving the relevant information in the data 40 . ...
... On the other hand, decreasing the cut-off parameter decreases the threshold values and leads to very sensitive filtering, causing not only the rejection of artifacts but also the loss of relevant information from the data 40 . Here, we adopted a conservative cut-off parameter of 30, which allows for the rejection of the majority of artifacts while preserving the relevant information in the data 40 . We visually inspected the data to further remove bad data periods not successfully repaired by the ASR 41 . ...
Previous studies have identified several brain regions involved in the sympathetic response and its integration with pain, cognition, emotions and memory processes. However, little is known about how such regions dynamically interact during a sympathetic activation task. In this study, we analyzed EEG activity and effective connectivity during a cold pressor test (CPT). A source localization analysis identified a network of common active sources including the right precuneus (r-PCu), right and left precentral gyri (r-PCG, l-PCG), left premotor cortex (l-PMC) and left anterior cingulate cortex (l-ACC). We comprehensively analyzed the network dynamics by estimating power variation and causal interactions among the network regions through the direct directed transfer function (dDTF). A connectivity pattern dominated by interactions in α (8–12) Hz band was observed in the resting state, with r-PCu acting as the main hub of information flow. After the CPT onset, we observed an abrupt suppression of such α-band interactions, followed by a partial recovery towards the end of the task. On the other hand, an increase of δ-band (1–4) Hz interactions characterized the first part of CPT task. These results provide novel information on the brain dynamics induced by sympathetic stimuli. Our findings suggest that the observed suppression of α and rise of δ dynamical interactions could reflect non-pain-specific arousal and attention-related response linked to stimulus’ salience.
... The drawback of ICA is that there needs to be a very large number of channels employed to collect the data relative to the underlying artifactual sources. To suppress the artifact from EEG data, the artifact subspace reconstruction (ASR) approach was developed in Chang et al. [23]. The effectiveness of the approach is controlled by the parameter. ...
... The effectiveness of the approach is controlled by the parameter. Despite doing a comprehensive analysis to identify the cut-off value, choosing it incorrectly might cause the EEG data to be lost [23,24]. ...
Background: Portable electroencephalogram (EEG) systems are often used in health care applications to record brain signals because their ease of use. An electrooculogram (EOG) is a common, low frequency, high amplitude artifact of the eye blink signal that might confuse disease diagnosis. As a result, artifact removal approaches in single EEG portable devices are in high demand. Materials: Dataset 2a from the BCI Competition IV was employed. It contains the EEG data from nine subjects. To determine the EOG effect, each session starts with 5 min of EEG data. This recording lasted for two minutes with the eyes open, one minute with the eyes closed, and one minute with eye movements. Methodology: This article presents the automated removal of EOG artifacts from EEG signals. Circulant Singular Spectrum Analysis (CiSSA) was used to decompose the EOG contaminated EEG signals into intrinsic mode functions (IMFs). Next, we identified the artifact signal components using kurtosis and energy values and removed them using 4-level discrete wavelet transform (DWT). Results: The proposed approach was evaluated on synthetic and real EEG data and found to be effective in eliminating EOG artifacts while maintaining low frequency EEG information. CiSSA-DWT achieved the best signal to artifact ratio (SAR), mean absolute error (MAE), relative root mean square error (RRMSE), and correlation coefficient (CC) of 1.4525, 0.0801, 18.274, and 0.9883, respectively. Comparison: The developed technique outperforms existing artifact suppression techniques according to performance measures. Conclusions: This advancement is important for brain science and can contribute as an initial pre-processing step for research related to EEG signals.
... Then, artifact subspace reconstruction (ASR) was performed using the default settings in EEGLab, followed by re-referencing to the global average. Here, ASR is an automated method based on a user-specified parameter that can effectively remove transient EEG artefacts (Chang et al., 2020). We used the default ASR parameter value of 20, and the optimal value is between 20 and 30 to balance between removing nonbrain signals and retaining brain activities (Chang et al., 2020). ...
... Here, ASR is an automated method based on a user-specified parameter that can effectively remove transient EEG artefacts (Chang et al., 2020). We used the default ASR parameter value of 20, and the optimal value is between 20 and 30 to balance between removing nonbrain signals and retaining brain activities (Chang et al., 2020). Then, a Laplacian spatial filter was applied to remove the volume conduction from the subcortical sources since we were interested in the cortical sources only that corresponded with the fNIRS HbO activity in our published work (Walia, Fu, Norfleet, et al., 2022). ...
The study objective was classification of skill level based on the topographical features of the electroencephalogram(EEG) during the most complex Fundamentals of Laparoscopic Surgery(FLS) task.
We developed a novel microstate-based Common Spatial Pattern (CSP) analysis with linear discriminant analysis(LDA) classification that was compared with topography-preserving convolutional neural network(CNN) based approach to distinguish experts versus novices based on EEG. Ten expert surgeons and thirteen novice medical residents were recruited at the University at Buffalo. After informed consent, the subjects performed three trials of laparoscopic suturing and knot tying with rest periods in-between. 32-channel EEG during task performance was used to analyze spatial patterns of brain activity in 8 expert surgeons (2 dropouts due to data quality) and 13 novice medical residents. Besides conventional CSP analysis, microstate analysis was applied for preprocessing before CSP analysis for improved classification using LDA with 10-fold cross-validation. Also, a topography-preserving 3D CNN model (ESNet) was applied that considered both spatial and temporal information for the classification. Here, 5-fold cross-validation was repeated 10 times, and the results of each iteration of the testing data set were evaluated using indices, Accuracy, F1 score, Mathews Correlation Coefficient (MCC), sensitivity, and Specificity.
Microstate-based CSP analysis found that while novices had primarily the frontal cortex involved for a maximum of spatial pattern vectors, experts had the hotspot of the spatial pattern vectors over the frontal and parietal cortices where the discriminating parietal brain region was supported by the Gradient-weighted Class Activation Mapping (Grad-CAM) of our 3D CNN-based model. Here, LDA with 10-fold cross-validation achieved more than 90% classification accuracy with microstate-based CSP, while conventional regularized CSP could reach around 80% classification accuracy. Then, 3D CNN provided the highest sensitivity of 99.30%, the highest specificity of 99.70%, the highest F1 score of 98.51%, and the highest MCC of 97.56%.
Microstate-based CSP analysis improved the LDA classification (~90%) of experts versus novices based on EEG topography during a complex FLS task; however, combining the spatial and temporal information in the EEG topography preserving 3D CNN model significantly improved the classifier accuracy (>98%) in addition to providing mechanistic insights based on Grad-CAM analysis.
... The clean_rawdata function also performed artifact subspace reconstruction (ASR). ASR is an automated method based on a user-specified parameter that can effectively remove transient EEG artifacts (Chang et al., 2020). We used a lower 'ChannelCriterion' parameter value of 0.8 while using other default parameters including 'BurstCriterion' ...
... parameter value of 20 where the optimal value is between 20 and 30 to balance between removing nonbrain signals and retaining brain activities (Chang et al., 2020). Also, the Picard algorithm (Frank et al., 2022) for Independent Component Analysis (ICA) pop_runica(ALLEEG, 'icatype','picard','concatcond','on','options',{'pca',-1}) eparate brain and non-brain related activities where the EEG in the same tDCS session was assumed by the clustering functions to have the same ICA component weights. ...
Transcranial direct current stimulation (tDCS) has been shown to facilitate surgical training and performance when compared to sham tDCS; however, the potency may be improved by selecting appropriate brain targets based on neuroimaging and mechanistic insights. Published studies have shown the feasibility of portable brain imaging in conjunction with tDCS during Fundamentals of Laparoscopic Surgery (FLS) tasks for concurrently monitoring the cortical activations via functional near-infrared spectroscopy (fNIRS). Then, fNIRS can be combined with electroencephalogram (EEG) where EEG band power changes have been shown to correspond to the changes in oxyhemoglobin (HbO) concentration, found from the fNIRS. In principal accordance with these prior works, our current study aimed to investigate multi-modal imaging of the brain response to cerebellar (CER) and ventrolateral prefrontal cortex (PFC) tDCS that may facilitate the most complex FLS suturing with intracorporal knot tying task. Our healthy human study on twelve novices (age: 22-28 years, 2 males, 1 female with left-hand dominance) from medical/premedical backgrounds aimed for mechanistic insights from neuroimaging brain areas that are related to error-based learning – one of the basic skill acquisition mechanisms. We found that right CER tDCS of the posterior lobe facilitated a statistically significant (q<0.05) brain response at the bilateral prefrontal areas at the start of the FLS task that was higher than sham tDCS. Also, right CER tDCS significantly (p<0.05) improved FLS score when compared to sham tDCS. In contrast, left PFC tDCS failed to facilitate a significant brain response and FLS performance improvement. Moreover, right CER tDCS facilitated activation of the bilateral prefrontal brain areas related to FLS performance improvement provided mechanistic insights into the CER tDCS effects. The mechanistic insights motivated future investigation of CER tDCS effects on the error-related perception action coupling based on directed functional connectivity studies.
... Artifact subspace reconstruction (ASR) was performed using the default settings in EEGlab, followed by re-referencing to the global average. ASR is an automated method based on a user-specified parameter that can effectively remove transient EEG artifacts [11]. We used the default ASR parameter value of 20, while the optimal value is between 20 and 30 to balance between removing non-brain signals and retaining brain activities [11]. ...
... ASR is an automated method based on a user-specified parameter that can effectively remove transient EEG artifacts [11]. We used the default ASR parameter value of 20, while the optimal value is between 20 and 30 to balance between removing non-brain signals and retaining brain activities [11]. The preprocessed EEG data from 13 novices and eight experts were used for the microstate analysis, since we rejected one expert subject to keep the maximum number of bad channels for any subject less than five. ...
Abstract Error-based learning is one of the basic skill acquisition mechanisms that can be modeled as a perception–action system and investigated based on brain–behavior analysis during skill training. Here, the error-related chain of mental processes is postulated to depend on the skill level leading to a difference in the contextual switching of the brain states on error commission. Therefore, the objective of this paper was to compare error-related brain states, measured with multi-modal portable brain imaging, between experts and novices during the Fundamentals of Laparoscopic Surgery (FLS) “suturing and intracorporeal knot-tying” task (FLS complex task)—the most difficult among the five psychomotor FLS tasks. The multi-modal portable brain imaging combined functional near-infrared spectroscopy (fNIRS) and electroencephalography (EEG) for brain–behavior analysis in thirteen right-handed novice medical students and nine expert surgeons. The brain state changes were defined by quasi-stable EEG scalp topography (called microstates) changes using 32-channel EEG data acquired at 250 Hz. Six microstate prototypes were identified from the combined EEG data from experts and novices during the FLS complex task that explained 77.14% of the global variance. Analysis of variance (ANOVA) found that the proportion of the total time spent in different microstates during the 10-s error epoch was significantly affected by the skill level (p
... respectively. The Graz University of Technology has many world-renowned scholars in the BCI field, including Pfurtscheller G, Neuper C, Muller-Putz GR, etc., which laid the foundation for its academic status of the highest h-index (76) in the realm of BCIs. A striking phenomenon is that the Wadsworth Center and University of New York (SUNY) system's ACPP is exceptionally high compared to other research institutions. ...
... However, the exemplary neural interface should integrate seamlessly with the nervous system and operate reliably for long periods. Developing such long-acting functional neural interface materials still faces many challenges [73][74][75][76][77], such as biocompatibility [105][106][107], mechanical mismatch [111][112][113][114][115], and electrical properties. observation suggests that fewer reviews have been made on ethical and nonmedical applications over the past decade, confirming that the current BCI research mainly relates to the rehabilitation application for the disabled. ...
Brain-computer interfaces (BCIs), as a promising technology for rehabilitation, have attracted significant interest in the fields such as neurosciences, computer science, and biomedical engineering. Based on the expanded database of Science Citation Index, this paper analyzes bibliometrics in the field of BCI from 1990 to 2020 and discusses its potential research trends and prospects. The work provides a detailed overview of the BCI research status founded on the country, research area, institution, journal, author, citation, and keyword. A total of 7,880 papers suggest that the United States is leading the field of BCI research, followed by China and Germany. University of California system has produced the most publications, while the Graz University of Technology is leading the list for the h-index. Journal of Neural Engineering and IEEE Transactions on Neural Systems and Rehabilitation Engineering were identified to be the most productive journals in BCIs. Keywords analysis reflects that most research has focused on electroencephalography (EEG)-based BCIs to achieve a safe, real-time, stable link between the patient and artificial actuators. We can conclude by analyzing hot articles that new materials applied to neural probes may become the next hot topic. The typical paradigm of the BCI industry is the most urgent problem that can be solved in the standardization process. The rehabilitation applications appear as the initial driving force and ultimate goal for BCIs. This research can offer valuable references to BCI practitioners and provide data support for related studies.
... Preprocessing methods included power spectral density and independent component analysis to detect the relative presence of high-and low-frequency activity (see Figure 1). Several types of filters were employed including traditional tunable high and low pass, common average reference spatial, and artifact subspace reconstruction (ASR; Chang et al., 2020). A traditional high-pass filter was set at 1 Hz and a low-pass filter set at 60 Hz were used to exclude artifacts above and below the frequencies of interest. ...
... Common average reference spatial filters calculated and removed the average signal from each of the eight channels. To further remove artifacts such as noise from electrical, environmental, and motion noise ASR filters were used to correct or remove transient or large-amplitude artifacts in data comprising multichannel EEG data (Chang et al., 2020). Principal component analysis (PCA) was repeatedly computed on covariance matrices. ...
Delirium occurs in as many as 80% of critically ill older adults and is associated with increased long-term cognitive impairment, institutionalization, and mortality. Less than half of delirium cases are identified using currently available subjective assessment tools. Electroencephalogram (EEG) has been identified as a reliable objective measure but has not been feasible. This study was a prospective pilot proof-of-concept study, to examine the use of machine learning methods evaluating the use of gamma band to predict delirium from EEG data derived from a limited lead rapid response handheld device. Data from 13 critically ill participants aged 50 or older requiring mechanical ventilation for more than 12 h were enrolled. Across the three models, accuracy of predicting delirium was 70 or greater. Stepwise discriminant analysis provided the best overall method. While additional research is needed to determine the best cut points and efficacy, use of a handheld limited lead rapid response EEG device capable of monitoring all five cerebral lobes of the brain for predicting delirium hold promise.
... A significant amount of research work has been conducted to remove noise and artifacts form EEG signals. Artifacts removal uses linear regression, filtering/regression, independent component analysis (ICA), and principal component analysis (PCA) or a combination of different techniques [21][22][23][24][25][26][27][28]. Currently, for regression or blind source separation, it is assumed that the EEG model is linear, whereas the noise that models the artifacts is additive. ...
... Quantitative EEG represents a set of features extracted from the EEG signals to assess the functional state of the brain. The frequency bands of clinical interest in which brain waves oscillate are delta (0.5-4 Hz), theta (4-7 Hz), alpha (8-13 Hz), beta (14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), and gamma (30-100 Hz) [30]. The features extraction uses signal-processingderived tools, such as signal powers, power spectrum parameters, regularity measures, and coherence [31][32][33][34]. ...
Traumatic brain injury (TBI) can produce temporary biochemical imbalance due to leaks through cell membranes or disruption of the axoplasmic flow due to the misalignment of intracellular neurofilaments. If untreated, TBI can lead to Alzheimer’s, Parkinson’s, or total disability. Mild TBI (mTBI) accounts for about about 90 percent of all TBI cases. The detection of TBI as soon as it happens is crucial for successful treatment management. Neuroimaging-based tests provide only a structural and functional mapping of the brain with poor temporal resolution. Such tests may not detect mTBI. On the other hand, the electroencephalogram (EEG) provides good spatial resolution and excellent temporal resolution of the brain activities beside its portability and low cost. The objective of this paper is to provide clinicians and scientists with a one-stop source of information to quickly learn about the different technologies used for TBI detection, their advantages and limitations. Our research led us to conclude that even though EEG-based TBI detection is potentially a powerful technology, it is currently not able to detect the presence of a mTBI with high confidence. The focus of the paper is to review existing approaches and provide the reason for the unsuccessful state of EEG-based detection of mTBI.
... Next, an independent component analysis was performed to remove eye movement artifacts by rejecting the independent components whose correlation coefficients with EOG signals were above 0.4 [37]. Artifact subspace reconstruction (ASR) was then applied to remove movement artifacts arisen from hand movement [38]. In this study, the cutoff parameter of ASR was set to be 30 as recommended in [38]. ...
... Artifact subspace reconstruction (ASR) was then applied to remove movement artifacts arisen from hand movement [38]. In this study, the cutoff parameter of ASR was set to be 30 as recommended in [38]. Next, the common average reference was used to remove spatial noise, and a zero-phase, fourth-order, low-pass Butterworth filter with a cutoff frequency of 2 Hz was used to extract low-frequency information from EEG signals [18] [19]. ...
The continuous decoding of human movement intention based on electroencephalogram (EEG) signals is valuable for developing a more natural motor augmented or assistive system instead of its discrete classifications. The classic center-out paradigm has been widely used to study discrete and continuous hand movement parameter decoding. However, when applying it in studying continuous movement decoding, the classic paradigm needs to be improved to increase the decoding performance, especially generalization performance. In this paper, we first discuss the limitations of the classic center-out paradigm in exploring the hand movement’s continuous decoding. Then, an improved paradigm is proposed to enhance the continuous decoding performance. Besides, an adaptive decoder-ensemble framework is developed for continuous kinematic parameter decoding. Finally, with the improved center-out paradigm and the ensemble decoding framework, the average Pearson’s correlation coefficients between the predicted and recorded movement kinematic parameters improve significantly by about 75 percent for the directional parameters and about 10 percent for the non-directional parameters. Furthermore, its generalization performance improves significantly by about 20 percent for the directional parameters. This study indicates the advantage of the improved paradigm in predicting the hand movement’s kinematic information from low-frequency scalp EEG signals. It can advance the applications of the noninvasive motor brain-computer interface (BCI) in rehabilitation, daily assistance, and human augmentation areas.
... [21], an interactive Matlab (MathWorks, Natick, MA, USA) toolbox was used for preprocessing the offline EEG data, e.g., data being downsampled to 500 Hz, band-pass filtered to 1-100 Hz with the finite impulse response method and analog 60 Hz-notch filtered. Bad channels were automatically detected and removed based on artifact subspace reconstruction (ASR) [22]. Independent component analysis (ICA) followed by ICLabel [23] was used to automatically remove artifacts caused by muscle activity, heartbeats, eye movements, and eye blinks (see Supplementary Materials). ...
EEG studies indicated that schizophrenia patients had increased resting-state theta-band functional connectivity, which was associated with negative symptoms. We recently published the first study showing that theta (6 Hz) transcranial alternating current stimulation (tACS) over left prefrontal and parietal cortices during a working memory task for accentuating frontoparietal theta-band synchronization (in-phase theta-tACS) reduced negative symptoms in schizophrenia patients. Here, we hypothesized that in-phase theta-tACS can modulate theta-band large-scale networks connectivity in schizophrenia patients. In this randomized, double-blind, sham-controlled trial, patients received twice-daily, 2 mA, 20-min sessions of in-phase theta-tACS for 5 consecutive weekdays (n = 18) or a sham stimulation (n = 18). Resting-state electroencephalography data were collected at baseline, end of stimulation, and at one-week follow-up. Exact low resolution electromagnetic tomography (eLORETA) was used to compute intra-cortical activity. Lagged phase synchronization (LPS) was used to measure whole-brain source-based functional connectivity across 84 cortical regions at theta frequency (5–7 Hz). EEG data from 35 patients were analyzed. We found that in-phase theta-tACS significantly reduced the LPS between the posterior cingulate (PC) and the parahippocampal gyrus (PHG) in the right hemisphere only at the end of stimulation relative to sham (p = 0.0009, corrected). The reduction in right hemispheric PC-PHG LPS was significantly correlated with negative symptom improvement at the end of the stimulation (r = 0.503, p = 0.039). Our findings suggest that in-phase theta-tACS can modulate theta-band large-scale functional connectivity pertaining to negative symptoms. Considering the failure of right hemispheric PC-PHG functional connectivity to predict improvement in negative symptoms at one-week follow-up, future studies should investigate whether it can serve as a surrogate of treatment response to theta-tACS.
... The ASR procedure was applied using a 500-ms sliding window and a lax (20 standard deviations) threshold that removes extreme mechanical artifacts while preserving brain signal components. This method has been shown to improve the quality of a subsequent Independent Component Analysis (ICA) decomposition [28,29]. Next, all removed channels were interpolated and EEG data were then re-referenced to a common average reference. ...
Background Is it possible to decode lying behavior from neural signatures of the electroencephalography (EEG)? Existing studies on this topic have methodological limitations: tasks lack the incentive to lie, the act of lying is confounded with memory recollection, there is no sufficient distinction between instructed versus spontaneous decisions, and participants' risk-taking tendency is not controlled for. To address these limitations, we introduce a novel interactive, two-player game, where successful lying is incentivized in the reward scheme and that has both instructed and spontaneous conditions for participants matched by risk-taking tendency.
Methods 24 participants were paired in the game according to their risk-taking tendency scores and measured using 32-channel EEG. Our multi-modal EEG analysis includes event-related potential (ERP), event-related spectral (ERS), and deep-learning-based single-trial decoding.
Results In ERP, two early components (P200 and N200) distinguished instructed truth-telling from other conditions, with a late component (N300) separating instructed lies from spontaneous conditions. Moreover, a late positive potential was indicative of spontaneous lying versus spontaneous truth-telling and was correlated with participants' risk-taking tendencies. In ERS, alpha and low beta were found to discriminate conditions. Importantly, we observed robust single-trial decoding performance under different conditions for the 1D-CNN. A gradient-based analysis further identified significant time-periods compatible with the ERP results from the classifier.
Conclusions Our study represents the first effort to analyze EEG data not only through statistical properties but also with out-of-sample, deep-learning-based classification to decode deceit in the context of an iterative, game-theoretic experiment.
... In [22], authors discussed the effectiveness of ASR and the optimal choice of its parameter. The authors presented an ASR evaluation study performed on twenty EEG recordings gathered during simulated driving trials. ...
EEG signals may be affected by physiological and non-physiological artifacts hindering the analysis of brain activity. Blind source separation methods such as independent component analysis (ICA) are effective ways of improving signal quality by removing components representing non-brain activity. However, most ICA-based artifact removal strategies have limitations, such as individual differences in visual assessment of components. These limitations might be reduced by introducing automatic selection methods for ICA components. On the other hand, new fully automatic artifact removal methods are developed. One of such method is artifact subspace reconstruction (ASR). ASR is a component-based approach, which can be used automatically and with small calculation requirements. The ASR was originally designed to be run not instead of, but in addition to ICA. We compared two automatic signal quality correction approaches: the approach based only on ICA method and the approach where ASR was applied additionally to ICA and run before the ICA. The case study was based on the analysis of data collected from 10 subjects performing four popular experimental paradigms, including resting-state, visual stimulation and oddball task. Statistical analysis of the signal-to-noise ratio showed a significant difference, but not between ICA and ASR followed by ICA. The results show that both methods provided a signal of similar quality, but they were characterised by different usabilities.
... Following data preprocessing, we primarily employed the power spectral density to quantify EEG data and the Fourier transform to derive the energy of the various EEG cycles. According to existing studies, the beta rhythm is chiefly associated with selective attention [35][36]. The subject's operational status might be indicated. ...
The advancement of virtual reality technology has ushered in new developments in the medical field. The use of virtual surgery training simulators alleviates the paucity of training resources and high training expenses associated with traditional surgical capabilities. Regardless of the type of schooling, doctors must continue to educate themselves. The postoperative evaluation mechanism is incomplete. Traditional objective evaluation indicators are unable to meet surgeons' stringent expectations. The EEG (Electroencephalogram) rhythm index is proposed in this article as a new tool for evaluating and distinguishing between novice and expert doctors. The experiment uses a cutting training module from neurosurgery training and compares it with established assessment metrics to determine the correct rate of classification of new evaluation metrics, classifying testers by both metrics and finding a 20% increase in correctness. Additionally, the paper compares the energy topographic maps of different EEG rhythms of novices and experts. For classification, two machine learning algorithms, SVM and random forest are utilized at the same time. The findings reveal that the accuracy of distinguishing indicators based on EEG cycles is 10% higher than that of typical objective evaluation indicators, regardless of the categorization method. ROC curve analysis was also used to compare the two classification models. The AUC value for the EEG rhythm evaluation index model was 0.971, whereas the AUC value for the classic objective evaluation index model was 0.761, which explains the EEG rhythm evaluation index. The model demonstrates a fairly reliable categorization standard.
... With large motion artifacts, there is a risk that ICA will not be able to extract high-quality brain components from the mixed data. Popular approaches for removing artifacts prior to ICA include adaptive filtering [19], principal component analysis [20,21], and wavelets [22]. Please see Seok et al. 2021 [23] for an in depth review of motion artifact removal methods for EEG. ...
Motion artifacts hinder source-level analysis of mobile electroencephalography (EEG) data using independent component analysis (ICA). iCanClean is a novel cleaning algorithm that uses reference noise recordings to remove noisy EEG subspaces, but it has not been formally tested in a parameter sweep. The goal of this study was to test iCanClean’s ability to improve the ICA decomposition of EEG data corrupted by walking motion artifacts. Our primary objective was to determine optimal settings and performance in a parameter sweep (varying the window length and r2 cleaning aggressiveness). High-density EEG was recorded with 120 + 120 (dual-layer) EEG electrodes in young adults, high-functioning older adults, and low-functioning older adults. EEG data were decomposed by ICA after basic preprocessing and iCanClean. Components well-localized as dipoles (residual variance < 15%) and with high brain probability (ICLabel > 50%) were marked as ‘good’. We determined iCanClean’s optimal window length and cleaning aggressiveness to be 4-s and r2 = 0.65 for our data. At these settings, iCanClean improved the average number of good components from 8.4 to 13.2 (+57%). Good performance could be maintained with reduced sets of noise channels (12.7, 12.2, and 12.0 good components for 64, 32, and 16 noise channels, respectively). Overall, iCanClean shows promise as an effective method to clean mobile EEG data.
... Line 60 Hz noises were eliminated using the EEGLAB plugin "CleanLine". High-amplitude artifacts were removed and reconstructed using the EEGLAB plugin "clean_rawdata()," including Artifact Subspace Reconstruction (ASR) [31][32][33][34][35][36][37]. For the removal of artifacts with this plugin, the following parameters were used: flat line removal, 5 s; electrode correlation, 0.9; ASR with correction criterion in SD, 15; window rejection with poor quality, 25%. ...
Gestures and speech, as linked communicative expressions, form an integrated system. Previous functional magnetic resonance imaging studies have suggested that neural networks for gesture and spoken word production share similar brain regions consisting of fronto-temporo-parietal brain regions. However, information flow within the neural network may dynamically change during the planning of two communicative expressions and also differ between them. To investigate dynamic information flow in the neural network during the planning of gesture and spoken word generation in this study, participants were presented with spatial images and were required to plan the generation of gestures or spoken words to represent the same spatial situations. The evoked potentials in response to spatial images were recorded to analyze the effective connectivity within the neural network. An independent component analysis of the evoked potentials indicated 12 clusters of independent components, the dipoles of which were located in the bilateral fronto-temporo-parietal brain regions and on the medial wall of the frontal and parietal lobes. Comparison of effective connectivity indicated that information flow from the right middle cingulate gyrus (MCG) to the left supplementary motor area (SMA) and from the left SMA to the left precentral area increased during gesture planning compared with that of word planning. Furthermore, information flow from the right MCG to the left superior frontal gyrus also increased during gesture planning compared with that of word planning. These results suggest that information flow to the brain regions for hand praxis is more strongly activated during gesture planning than during word planning.
... The movement of the eyelid creates ocular artifacts and these artifacts are mostly affected by lateral frontal electrodes [6]. Various algorithms are developed for the removal of ocular artifacts [7][8][9][10][11][12][13][14]. The ECG artifacts are mostly observed at mid-temporal and posterior temporal electrodes. ...
... Additionally, before FTA processing, data are preprocessed using the Artifacts Subspace Reconstruction (ASR) algorithm [27] (as we observed poor performance in SSVEP detection using the raw filtered data). For this, the default settings (k = 20, ASR Removal) are used, as they empirically yield better results [12,28,29]. Trials with a duration of less than 20 s were discarded from further analyses. ...
Recent studies show that the integrity of core perceptual and cognitive functions may be tested in a short time with Steady-State Visual Evoked Potentials (SSVEP) with low stimulation frequencies, between 1 and 10 Hz. Wearable EEG systems provide unique opportunities to test these brain functions on diverse populations in out-of-the-lab conditions. However, they also pose significant challenges as the number of EEG channels is typically limited, and the recording conditions might induce high noise levels, particularly for low frequencies. Here we tested the performance of Normalized Canonical Correlation Analysis (NCCA), a frequency-normalized version of CCA, to quantify SSVEP from wearable EEG data with stimulation frequencies ranging from 1 to 10 Hz. We validated NCCA on data collected with an 8-channel wearable wireless EEG system based on BioWolf, a compact, ultra-light, ultra-low-power recording platform. The results show that NCCA correctly and rapidly detects SSVEP at the stimulation frequency within a few cycles of stimulation, even at the lowest frequency (4 s recordings are sufficient for a stimulation frequency of 1 Hz), outperforming a state-of-the-art normalized power spectral measure. Importantly, no preliminary artifact correction or channel selection was required. Potential applications of these results to research and clinical studies are discussed.
... Remaining channels were then visualized, and additional noisy channels were removed manually after inspection by several trained research assistants (average number of channels removed per participant = 3.45, SD = 3.10). Following the removal of bad channels, non-linearities in the electroencephalogram were corrected using the artifact subspace reconstruction (ASR) algorithm [77] with a standard deviation threshold set to 20 and k-window set to 0.25-parameters which have been shown to correct artifact-driven non-linearities in EEG data, while preserving brain-related activity [78]. Finally, EEG source decomposition was conducted, using independent components analysis, to remove components that reflected eye and cranial-muscle artifacts. ...
Neural entrainment to musical rhythm is thought to underlie the perception and production of music. In aging populations, the strength of neural entrainment to rhythm has been found to be attenuated, particularly during attentive listening to auditory streams. However, previous studies on neural entrainment to rhythm and aging have often employed artificial auditory rhythms or limited pieces of recorded, naturalistic music, failing to account for the diversity of rhythmic structures found in natural music. As part of larger project assessing a novel music-based intervention for healthy aging, we investigated neural entrainment to musical rhythms in the electroencephalogram (EEG) while participants listened to self-selected musical recordings across a sample of younger and older adults. We specifically measured neural entrainment to the level of musical pulse—quantified here as the phase-locking value (PLV)—after normalizing the PLVs to each musical recording’s detected pulse frequency. As predicted, we observed strong neural phase-locking to musical pulse, and to the sub-harmonic and harmonic levels of musical meter. Overall, PLVs were not significantly different between older and younger adults. This preserved neural entrainment to musical pulse and rhythm could support the design of music-based interventions that aim to modulate endogenous brain activity via self-selected music for healthy cognitive aging.
... It adopts a sliding window (default 0.5 s, overlap 50%) to principal component analysis (PCA) and decompose all the channels. This method identifies 'bad' Principal components (PC's) (defined by comparison against the data's own cleanest part in frequency-enhanced root mean square) to reject and reconstruct activity from the remaining components [85]. ASR process consists of three steps: (1) extracting reference data from raw data, (2) determining thresholds for identifying artifact components, and (3) After cleaning the dataset with ASR, dataset is segmented in 4 seconds intervals; it is a suitable duration for effective connectivity analysis using multivariate autoregressive models in which the input time series should be stable and stationary. ...
Abstract
Effective connectivity analysis promotes knowledge of cognitive processes in the brain
by modeling the cause-and-effect interactions among the areas. This preliminary study
compares the directional coefficients of the default mode network among six stroke
patients and nine healthy participants. Furthermore, the alterations of connections are
studied after rehabilitation in this project. This information leads to a better
understanding of brain aberrations after stroke and supports a better rehabilitation
program. This is one of the few study that examines effective connectivity in stroke
cases using EEG signals. EEG provides a higher accuracy by maintaining a better
temporal precision (<1 ms) necessary for estimating neural activities over time.
The outcome of this study confirms the results of previous publications; this project
also improves the accuracy and determines the direction of the connections compared
to former studies. Significant differences are recognized in the links toward the
posterior cingulate within stroke and healthy groups and pre- and post-rehabilitation
structures. This result reinforces the importance of this region in stroke studies.
Reduced connections are observed from anterior cingulate to Dorsolateral Prefrontal
cortex (Right) and from Dorsolateral Prefrontal cortex(left) to Inferior Parietal Lobule
(left) in stroke cases compared to healthy ones. The rehabilitation process restores
these connections. The connection from the Dorsolateral Prefrontal cortex to the
Inferior Parietal Lobule is also improved during the recovery. This study endorses that
the rehabilitation process reconstructs the altered relations in stroke cases.
Key-words: effective Connectivity, stroke, EEG, eLORETA, isolated effective
coherence, default mode network.
... The sampled raw EEG signals were processed through the following procedure: (1) High-pass filtering above 0.5 Hz was applied; (2) For each channel, if more than 70% of all other channels exhibited a cross-correlation lower than 0.4 with that channel after band-pass filtering between 0.5 and 1 Hz, the channel was deemed as a bad one and was removed (Bigdely-Shamlo et al. 2015); (3) Potential noise components from the reference were removed using the common average reference (CAR) technique; (4) The re-referenced signals were low-pass filtered below 50 Hz; (5) Artifacts were minimized using the artifact subspace reconstruction (ASR) method (Bigdely-Shamlo et al. 2015;Chang et al. 2020;Mullen et al. 2015); and (6) Low-pass filtering below 12 Hz was applied to the signals for the ERP analysis. The average number of bad channels was two across all the BCI systems. ...
Non-invasive brain-computer interfaces (BCIs) based on an event-related potential (ERP) component, P300, elicited via the oddball paradigm, have been extensively developed to enable device control and communication. While most P300-based BCIs employ visual stimuli in the oddball paradigm, auditory P300-based BCIs also need to be developed for users with unreliable gaze control or limited visual processing. Specifically, auditory BCIs without additional visual support or multi-channel sound sources can broaden the application areas of BCIs. This study aimed to design optimal stimuli for auditory BCIs among artificial (e.g., beep) and natural (e.g., human voice and animal sounds) sounds in such circumstances. In addition, it aimed to investigate differences between auditory and visual stimulations for online P300-based BCIs. As a result, natural sounds led to both higher online BCI performance and larger differences in ERP amplitudes between the target and non-target compared to artificial sounds. However, no single type of sound offered the best performance for all subjects; rather, each subject indicated different preferences between the human voice and animal sound. In line with previous reports, visual stimuli yielded higher BCI performance (average 77.56%) than auditory counterparts (average 54.67%). In addition, spatiotemporal patterns of the differences in ERP amplitudes between target and non-target were more dynamic with visual stimuli than with auditory stimuli. The results suggest that selecting a natural auditory stimulus optimal for individual users as well as making differences in ERP amplitudes between target and non-target stimuli more dynamic may further improve auditory P300-based BCIs.
... EEGLab toolbox version 2021.0 from Matlab was used to pre-process and process the data. High variance spontaneous artifacts were removed by the Artifact Subspace Reconstruction algorithm (Chang et al., 2020). Bad channels were depicted as (1) having a flatline longer than 5 s, or (2) presenting more line noise relative to its signal than 4 standard deviations based on the total channel population, or (3) channels which joint 10.3389/fncom.2022.1022787 ...
Artificial voices are nowadays embedded into our daily lives with latest neural voices approaching human voice consistency (naturalness). Nevertheless, behavioral, and neuronal correlates of the perception of less naturalistic emotional prosodies are still misunderstood. In this study, we explored the acoustic tendencies that define naturalness from human to synthesized voices. Then, we created naturalness-reduced emotional utterances by acoustic editions of human voices. Finally, we used Event-Related Potentials (ERP) to assess the time dynamics of emotional integration when listening to both human and synthesized voices in a healthy adult sample. Additionally, listeners rated their perceptions for valence, arousal, discrete emotions, naturalness, and intelligibility. Synthesized voices were characterized by less lexical stress (i.e., reduced difference between stressed and unstressed syllables within words) as regards duration and median pitch modulations. Besides, spectral content was attenuated toward lower F2 and F3 frequencies and lower intensities for harmonics 1 and 4. Both psychometric and neuronal correlates were sensitive to naturalness reduction. (1) Naturalness and intelligibility ratings dropped with emotional utterances synthetization, (2) Discrete emotion recognition was impaired as naturalness declined, consistent with P200 and Late Positive Potentials (LPP) being less sensitive to emotional differentiation at lower naturalness, and (3) Relative P200 and LPP amplitudes between prosodies were modulated by synthetization. Nevertheless, (4) Valence and arousal perceptions were preserved at lower naturalness, (5) Valence (arousal) ratings correlated negatively (positively) with Higuchi's fractal dimension extracted on neuronal data under all naturalness perturbations, (6) Inter-Trial Phase Coherence (ITPC) and standard deviation measurements revealed high inter-individual heterogeneity for emotion perception that is still preserved as naturalness reduces. Notably, partial between-participant synchrony (low ITPC), along with high amplitude dispersion on ERPs at both early and late stages emphasized miscellaneous emotional responses among subjects. In this study, we highlighted for the first time both behavioral and neuronal basis of emotional perception under acoustic naturalness alterations. Partial dependencies between ecological relevance and emotion understanding outlined the modulation but not the annihilation of emotional integration by synthetization.
... Next, the EEG data were re-referenced to the averaged earlobes (i.e., A1 and A2) and the common average of all channels successively. The motion-related artifacts were removed by the artifact subspace reconstruction (ASR) algorithm [30]. Additionally, eye blinks and eye movement artifacts were visually identified and removed from EEG using the independent component analysis (ICA) method [31]. ...
Bimanual coordination is common in human daily life, whereas current research focused mainly on decoding unimanual movement from electroencephalogram (EEG) signals. Here we developed a brain-computer interface (BCI) paradigm of task-oriented bimanual movements to decode coordinated directions from movement-related cortical potentials (MRCPs) of EEG. Eight healthy subjects participated in the target-reaching task, including (1) performing leftward, midward, and rightward bimanual movements, and (2) performing leftward and rightward unimanual movements. A combined deep learning model of convolution neural network and bidirectional long short-term memory network was proposed to classify movement directions from EEG. Results showed that the average peak classification accuracy for three coordinated directions of bimanual movements reached 73.39 ± 6.35%. The binary classification accuracies achieved 80.24 ± 6.25, 82.62 ± 7.82, and 86.28 ± 5.50% for leftward versus midward, rightward versus midward and leftward versus rightward, respectively. We also compared the binary classification (leftward versus rightward) of bimanual, left-hand, and right-hand movements, and accuracies achieved 86.28 ± 5.50%, 75.67 ± 7.18%, and 77.79 ± 5.65%, respectively. The results indicated the feasibility of decoding human coordinated directions of task-oriented bimanual movements from EEG.
... The noiseless signals are reconstructed by preserving the components without carrying artifacts and back-projected to the time domain. The ASR method has been shown capable of improving the quality of ICA decomposition (Chang et al., 2020). ...
Human electroencephalography (EEG) is a brain monitoring modality that senses cortical neuroelectrophysiological activity in high-temporal resolution. One of the greatest challenges posed in applications of EEG is the unstable signal quality susceptible to inevitable artifacts during recordings. To date, most existing techniques for EEG artifact removal and reconstruction are applicable to offline analysis solely, or require individualized training data to facilitate online reconstruction. We have proposed CLEEGN, a novel convolutional neural network for plug-and-play automatic EEG reconstruction. CLEEGN is based on a subject-independent pre-trained model using existing data and can operate on a new user without any further calibration. The performance of CLEEGN was validated using multiple evaluations including waveform observation, reconstruction error assessment, and decoding accuracy on well-studied labeled datasets. The results of simulated online validation suggest that, even without any calibration, CLEEGN can largely preserve inherent brain activity and outperforms leading online/offline artifact removal methods in the decoding accuracy of reconstructed EEG data. In addition, visualization of model parameters and latent features exhibit the model behavior and reveal explainable insights related to existing knowledge of neuroscience. We foresee pervasive applications of CLEEGN in prospective works of online plug-and-play EEG decoding and analysis.
... Because of the known technical issues with the Sedline ® EEG export [40], we down-sampled every recording to 89 Hz and only considered the 2nd channels displayed centrally (R1), as the other channels displayed at the top and the bottom were at higher risk to be affected by clipping. We then performed an artifact subspace reconstruction with the functions from the EEGLAB toolbox (Swartz Center for Computational Neuroscience, Institute for Neural Computation, University of California San Diego, La Jolla, CA 92093-0961, USA) [41], and the burst criterion was set to 20 as recommended [42]. We then calculated the DSA using the Welch functions with default settings and NFFT set to 256 (i.e., a frequency resolution set to 0.35 Hz). ...
This study aimed to investigate the use of electroencephalography (EEG) for detecting brain activity changes perioperatively in anesthetized horses subjected to surgery. Twelve adult horses undergoing various surgeries were evaluated after premedication with xylazine and butorphanol, induction with ketamine, midazolam, and guaifenesin, and maintenance with isoflurane. The frontal EEG electrodes were placed after the horse was intubated and mechanically ventilated. The EEG data were collected continuously from Stage (S)1—transition from induction to isoflurane maintenance, S2—during surgery, S3—early recovery before xylazine sedation (0.2 mg kg IV), and S4—recovery after xylazine sedation. The Patient State Index (PSI), (Burst) Suppression Ratio (SR), and 95% Spectral Edge Frequency (SEF95) were compared across the stages. The PSI was lowest in S2 (20.8 ± 2.6) and increased to 30.0 ± 27.7 (p = 0.005) in S3. The SR increased from S1 (5.5 ± 10.7%) to S3 (32.7 ± 33.8%, p = 0.0001). The spectral power analysis showed that S3 had a significantly higher content of delta wave activity (0.1–4 Hz) in the EEG and lower relative power in the 3 Hz to 15 Hz range when compared to S1 and S2. A similar result was observed in S4, but the lower power was in a narrower range, from 3 Hz to 7 Hz, which indicate profound central nervous system depression potentiated by xylazine, despite the cessation of isoflurane anesthesia. We concluded that the use of EEG provides clinically relevant information about perioperative brain state changes of the isoflurane-anesthetized horse.
... [23], an interactive Matlab (MathWorks, Natick, MA, USA) toolbox. Bad channels were automatically detected and removed based on artifact subspace reconstruction (ASR) [24]. Independent component analysis (ICA) followed by ICLabel [25] was used to automatically remove artifacts caused by muscle activity, heartbeats, eye movements, and eye blinks. ...
Schizophrenia is associated with increased resting-state large-scale functional network connectivity in the gamma frequency. High-frequency transcranial random noise stimulation (hf-tRNS) modulates gamma-band endogenous neural oscillations in healthy individuals through the application of low-amplitude electrical noises. Yet, it is unclear if hf-tRNS can modulate gamma-band functional connectivity in patients with schizophrenia. We performed a randomized, double-blind, sham-controlled clinical trial to contrast hf-tRNS (N = 17) and sham stimulation (N = 18) for treating negative symptoms in 35 schizophrenia patients. Short continuous currents without neuromodulatory effects were applied in the sham group to mimic real-stimulation sensations. We used electroencephalography to investigate if a five-day, twice-daily hf-tRNS protocol modulates gamma-band (33–45 Hz) functional network connectivity in schizophrenia. Exact low resolution electromagnetic tomography (eLORETA) was used to compute intra-cortical activity from regions within the default mode network (DMN) and fronto-parietal network (FPN), and functional connectivity was computed using lagged phase synchronization. We found that hf-tRNS reduced gamma-band within-DMN and within-FPN connectivity at the end of stimulation relative to sham stimulation. A trend was obtained between the change in within-FPN functional connectivity from baseline to the end of stimulation and the improvement of negative symptoms at the one-month follow-up (r = −0.49, p = 0.055). Together, our findings suggest that hf-tRNS has potential as a network-level approach to modulate large-scale functional network connectivity pertaining to negative symptoms of schizophrenia.
... an open-source toolbox for signal processing [28]. The artifact subspace reconstruction (ASR) method was applied to automatically detect and remove the bad channels [29]. Independent component analysis (ICA) was used to separate a multivariate signal into additive subcomponents and ICLabel [30] was applied to remove artifacts caused by muscle activity, heartbeats, eye movements, and eye blinks. ...
Reduced left-lateralized electroencephalographic (EEG) frontal alpha asymmetry (FAA), a biomarker for the imbalance of interhemispheric frontal activity and motivational disturbances, represents a neuropathological attribute of negative symptoms of schizophrenia. Unidirectional high-frequency transcranial random noise stimulation (hf-tRNS) can increase the excitability of the cortex beneath the stimulating electrode. Yet, it is unclear if hf-tRNS can modulate electroencephalographic FAA in patients with schizophrenia. We performed a randomized, double-blind, sham-controlled clinical trial to contrast hf-tRNS and sham stimulation for treating negative symptoms in 35 schizophrenia patients. We used electroencephalography to investigate if 10 sessions of hf-tRNS delivered twice-a-day for five consecutive weekdays would modulate electroencephalographic FAA in schizophrenia. EEG data were collected and FAA was expressed as the differences between common-log-transformed absolute power values of frontal right and left hemisphere electrodes in the alpha frequency range (8–12.5 Hz). We found that hf-tRNS significantly increased FAA during the first session of stimulation (p = 0.009) and at the 1-week follow-up (p = 0.004) relative to sham stimulation. However, FAA failed to predict and surrogate the improvement in the severity of negative symptoms with hf-tRNS intervention. Together, our findings suggest that modulating electroencephalographic frontal alpha asymmetry by using unidirectional hf-tRNS may play a key role in reducing negative symptoms in patients with schizophrenia.
... Channels containing artifacts were identified using the clean_artifacts function using the following 3 parameters: (a) channel flatline lasting longer than five seconds, (b) channel noise exceeding four standard deviations relative to its own signal and (c) a correlation < 0.85 with its neighboring channels. General data cleaning was performed using the validated Artifact Subspace Reconstruction (ASR) method 62 . ASR decomposes the timeseries data into principal components and detects artifactual components by comparing specific components with components from the data's cleanest segments. ...
Major Depressive Disorder (MDD) is a widespread mental illness that causes considerable suffering, and neuroimaging studies are trying to reduce this burden by developing biomarkers that can facilitate detection. Prior fMRI- and neurostimulation studies suggest that aberrant subgenual Anterior Cingulate (sgACC)—dorsolateral Prefrontal Cortex (DLPFC) functional connectivity is consistently present within MDD. Combining the need for reliable depression markers with the electroencephalogram’s (EEG) high clinical utility, we investigated whether aberrant EEG sgACC–DLPFC functional connectivity could serve as a marker for depression. Source-space Amplitude Envelope Correlations (AEC) of 20 MDD patients and 20 matched controls were contrasted using non-parametric permutation tests. In addition, extracted AEC values were used to (a) correlate with characteristics of depression and (b) train a Support Vector Machine (SVM) to determine sgACC–DLPFC connectivity’s discriminative power. FDR-thresholded statistical maps showed reduced sgACC–DLPFC AEC connectivity in MDD patients relative to controls. This diminished AEC connectivity is located in the beta-1 (13–17 Hz) band and is associated with patients’ lifetime number of depressive episodes. Using extracted sgACC–DLPFC AEC values, the SVM achieved a classification accuracy of 84.6% (80% sensitivity and 89.5% specificity) indicating that EEG sgACC–DLPFC connectivity has promise as a biomarker for MDD.
... Then, data were 1 Hz high-pass filtered (Klug & Gramann, 2021) using pop_eegfiltnew function, and 50 Hz line noise was reduced using the ZapLine plugin (de Cheveigné, 2020). We used Artifact Substance Reconstruction (ASR) to clean continuous EEG using a cutoff parameter at 20 (Chang et al., 2020). We then average-referenced the data and ran Independent Component Analysis (ICA; extended infomax algorithm). ...
Learning to pronounce a foreign phoneme requires an individual to acquire a motor program that enables the reproduction of the new acoustic target sound. This process is largely based on the use of auditory feedback to detect pronunciation errors to adjust vocalization. While early auditory evoked neural activity underlies automatic detection and adaptation to vocalization errors, little is known about the neural correlates of acquiring novel speech targets. To investigate the neural processes that mediate the learning of foreign phoneme pronunciation, we recorded event-related potentials (ERP) when participants (N=19) pronounced native or foreign phonemes. Behavioral results indicated that the participants’ pronunciation of the foreign phoneme improved during the experiment. Early auditory responses (N1 and P2 waves, approx. 85–290 ms after the sound onset) revealed no differences between foreign and native phonemes. In contrast, the amplitude of the fronto-centrally distributed late slow wave (LSW, 320–440 ms) was modulated by the pronunciation of the foreign phonemes, and the effect changed during the experiment, paralleling the improvement in pronunciation. These results suggest that the LSW may reflect higher-order monitoring processes that signal successful pronunciation and help learn novel phonemes.
... In our analysis, the resulting filter order was 1690 for the high pass filter and 170 for the low pass filter. Abrupt artifacts were corrected using the Artifact Subspace Reconstruction method (function clean_asr with a cut-off value of 20 SD; see Chang et al., 2020). Independent Component Analysis (ICA) was performed to correct for eye movements (Delorme and Makeig, 2004). ...
In meditation practices that involve focused attention to a specific object, novice practitioners often experience moments of distraction (i.e., mind wandering). Previous studies have investigated the neural correlates of mind wandering during meditation practice through Electroencephalography (EEG) using linear metrics (e.g., oscillatory power). However, their results are not fully consistent. Since the brain is known to be a chaotic/nonlinear system, it is possible that linear metrics cannot fully capture complex dynamics present in the EEG signal. In this study, we assess whether nonlinear EEG signatures can be used to characterize mind wandering during breath focus meditation in novice practitioners. For that purpose, we adopted an experience sampling paradigm in which 25 participants were iteratively interrupted during meditation practice to report whether they were focusing on the breath or thinking about something else. We compared the complexity of EEG signals during mind wandering and breath focus states using three different algorithms: Higuchi's fractal dimension (HFD), Lempel-Ziv complexity (LZC), and Sample entropy (SampEn). Our results showed that EEG complexity was generally reduced during mind wandering relative to breath focus states. We conclude that EEG complexity metrics are appropriate to disentangle mind wandering from breath focus states in novice meditation practitioners, and therefore, they could be used in future EEG neurofeedback protocols to facilitate meditation practice.
... But especially in mobile experiments, removing time periods with muscle activity would result in excessive cleaning, and the computed ICA decomposition would not be readily applicable to the entire data set since it was not informed by time points containing muscle activity. And while the cleaning threshold of ASR can be adjusted to remove mainly large transient spikes, ASR is very sensitive to this threshold (Chang et al., 2020), and especially for mobile data, it does not always find a suitable baseline by itself. ASR thus requires a specifically recorded baseline and sometimes different cleaning thresholds for different movement modalities and even different data sets within the same modality, which renders it unsuitable for automatic data cleaning as targeted in this study. ...
Objective. Electroencephalography (EEG) studies increasingly make use of more ecologically valid experimental protocols involving mobile participants that actively engage with their environment leading to increased artifacts in the recorded data (MoBI; Gramann et al., 2011). When analyzing EEG data, especially in the mobile context, removing samples regarded as artifactual is a common approach before computing independent component analysis (ICA). Automatic tools for this exist, such as the automatic sample rejection of the AMICA algorithm (Palmer et al., 2011), but the impact of both movement intensity and the automatic sample rejection has not been systematically evaluated yet.
Approach. We computed AMICA decompositions on eight data sets from six open-access studies with varying degrees of movement intensities using increasingly conservative sample rejection criteria. We evaluated the subsequent decomposition quality in terms of the component mutual information, the amount of brain, muscle, and "other" components, the residual variance of the brain components, and an exemplary signal-to-noise ratio.
Main results. We found that increasing movements of participants led to decreasing decomposition quality for individual data sets but not as a general trend across all movement intensities. The cleaning strength had less impact on decomposition results than anticipated, and moderate cleaning of the data resulted in the best decompositions.
Significance. Our results indicate that the AMICA algorithm is very robust even with limited data cleaning. Moderate amounts of cleaning such as 5 to 10 iterations of the AMICA sample rejection with 3 standard deviations as the threshold will likely improve the decomposition of most data sets, irrespective of the movement intensity.
... Second, each EEG time series was re-referenced to the voltage average over all the 19 channels. Next, to remove noise and artifacts [71,72] such as eye movements, saccades, and jaw clenching, independent component analysis (ICA) [73,74] was used, and finally, artifact subspace reconstruction (ASR) was used to reconstruct bad epochs using principal component analysis (PCA) [182]. Then, clean EEG signals from electrodes Fp1 and Fp2 were selected and band-pass filtered in the 13-30 Hz range to obtain beta-band temporal activity in the prefrontal area. ...
Transcranial photobiomodulation (tPBM) targets the human brain with near-infrared (NIR) light and is shown to affect human cognitive performance and neural electrophysiological activity as well as concentration changes of oxidized cytochrome-c-oxidase ([CCO]) and hemoglobin oxygenation ([HbO]) in human brain.
Brain topographical connectivity, which shows the communication between regions of the brain, and its alteration can be assessed to quantify the effects of external stimuli, diseases, and cognitive decline, in resting-state or task-based measurements. Furthermore, understanding the interactions between different physiological representations of neural activity, namely electrophysiological, hemodynamic, and metabolic signals in the human brain, has been an important topic among researchers in recent decades. In my doctoral study, neurophysiological networks were constructed using frequency-domain analyses on oscillations of electroencephalogram (EEG), [CCO], and [HbO] time series that were acquired by a portable EEG and 2-channel broadband near-infrared spectroscopy (2-bbNIRS).
Specifically, my dissertation included three aims. The first one was to examine how tPBM altered the topographical connectivity in the electrophysiological oscillations of the resting human brain. As the first step, I defined and found key regions and clusters in the EEG sensor space that were affected the most by tPBM during and after the stimulation using both cluster-based power analysis and graph-based connectivity analysis. The results showed that the right prefrontal 1064-nm tPBM modulates several global and regional electrophysiological networks by shifting the information path towards frontal regions, especially in the beta band. For the second aim, I performed 2-bbNIRS measurements from 26 healthy humans and developed a methodology that enabled quantification of the infra-slow oscillation (ISO) power and connectivity between bilateral frontal regions of the human brain in resting state and in response to frontal tPBM stimulation at different sites and laser wavelengths. As the result, several stable and consistent features were extracted in the resting state of 26 young healthy adults. Moreover, these features were used to reveal some effects of tPBM on prefrontal metabolism and hemodynamics, while illustrating the similarities and differences between different stimulation conditions. Finally, the third aim was to investigate the resting-state prefrontal physiological network and the corresponding modulation in response to left frontal 800-nm tPBM by determining the effective connectivity/coupling between each pair of the electrophysiological, hemodynamic, and metabolic ISO of the human brain. Complementary to the previous studies, my study showed that prefrontal tPBM not only modulates the information path between two locations of the prefrontal cortex, it can also induce unilateral alterations in interactions between neural activity, hemodynamics, and metabolism. Overall, my dissertation shed light on the mechanism of action of prefrontal tPBM.
... power line inference) noise. The Artefact Subspace Reconstruction (ASR) with a default cut-off parameter (k = 20) was applied to remove non-stationary artefacts [18]. The data were then decomposed into Independent Components (ICs), using the SOBI algorithm [19]. ...
The ability to manage the emotions has been associated to the Emotional Styles (ES), a set of coherent ways to deal with life's experiences. Recently, the Emotional Style Questionnaire (ESQ) has been proposed as a self-report mea-sure to assess the individual ES. The present study investigates the spectral differences in the resting-state EEG due to the individual ES, in order to support the psychometric reliability of the ESQ with associated neurophysiological measurements. In the alpha and beta band, Social Intuition showed significant and large (d > 0.8) effect sizes on the parietal and parieto-occipital regions, as well as a significant and large effect size in the gamma band on the pre-frontal region. In the beta band, Attention showed a significant and large effect size on the parieto-occipital region.
Objective:
Electroencephalographic (EEG) data are often contaminated with non-neural artifacts which can confound experimental results. Current artifact cleaning approaches often require costly manual input. Our aim was to provide a fully automated EEG cleaning pipeline that addresses all artifact types and improves measurement of EEG outcomes METHODS: We developed RELAX (the Reduction of Electroencephalographic Artifacts). RELAX cleans continuous data using Multi-channel Wiener filtering [MWF] and/or wavelet enhanced independent component analysis [wICA] applied to artifacts identified by ICLabel [wICA_ICLabel]). Several versions of RELAX were compared using three datasets (N = 213, 60 and 23 respectively) against six commonly used pipelines across a range of artifact cleaning metrics, including measures of remaining blink and muscle activity, and the variance explained by experimental manipulations after cleaning.
Results:
RELAX with MWF and wICA_ICLabel showed amongst the best performance at cleaning blink and muscle artifacts while preserving neural signal. RELAX with wICA_ICLabel only may perform better at differentiating alpha oscillations between working memory conditions.
Conclusions:
RELAX provides automated, objective and high-performing EEG cleaning, is easy to use, and freely available on GitHub.
Significance:
We recommend RELAX for data cleaning across EEG studies to reduce artifact confounds, improve outcome measurement and improve inter-study consistency.
Tinnitus is an auditory condition that causes humans to hear a sound at any time anywhere. Chronic and refractory tinnitus is caused by the over-synchronization of neurons, and a treatment alternative has been created on the basis of the re-synchronization of neural activity by means of sound. To date, several acoustic therapies have been proposed to treat tinnitus. However, the effect is not well understood yet. Therefore, the aim of this study is to evaluate the effect of auditory discrimination therapy (ADT) by monitoring the level of neural synchronization before and after the ADT-based treatment. To this end, we mapped and analyzed the neural (de-) synchronization response of 11 tinnitus patients in two experimental conditions: (1) auditory material encoding and (2) auditory material retrieval. The results suggest that event-related (de-) synchronization (ERD/ERS) maps could indicate ADT reduces attention toward tinnitus if we consider incremental alpha-ERS responses elicited after the acoustic therapy during auditory encoding tasks. Furthermore, there were significant statistical changes in at least one of the brain frequency bands in each subject, which could indicate a decrease in attention headed for tinnitus or toward everyday acoustic environments. Furthermore, a cross-sectional study, comparing the patients with regard to the control subjects showed up that tinnitus sufferers were able to identify typical related auditory stimulus with a high complexity level alongside with an increase in their cognitive demands.
Feeding assistance is one of the most fundamental and time-consuming activities of daily living. In this study, we designed, implemented, and tested a system for virtual response testing of a human-in-the-loop assistive feeding robot with physiological measurements. The study focused on how different levels of autonomy (fully autonomous vs. partial autonomous) in feeding robots affected user experience. In a within-subject experiment with randomized order, we found statistically significant differences in Duration, Usability, Workload, and Success rate between Autonomy Modes. The results from EEG measures were consistent with the self-reported results in evaluating the Usability and Workload. Our findings support the potential of VR simulation and biometric sensors as an effective way to evaluate user interactions with an assistive robot, and suggest users’ preference to collaborate with the robot.
Understanding the hotspots and patterns of tourism poverty alleviation is of great significance to stakeholders. This study aims to comprehensively and scientifically understand the research hotspots, characteristics, trends, and future research directions in the field of tourism poverty alleviation, provide decision support for tourism stakeholders, improve the level of management and service quality of tourism, balance the distribution of tourism resources and interests, and promote high‐quality and sustainable development of tourism. First, we propose a dynamic nonnegative matrix factorization (DNMF) method to analyze the tourism poverty alleviation literature from 2012 to 2021, extract hot topics and topic words. Then, we propose a topic evaluation criteria to evaluate the effect of the extracted topics. Third, we propose a topic intensity model to calculate the intensity of hot topics that can track the evolution trend of hot topics in the past 10 years, and explore the future development trends of the tourism poverty alleviation. Finally, we put forward the countermeasures and suggestions for sustainable tourism poverty alleviation under the guidance of sustainable development theory. The experimental results show that DNMF has a good effect in topic extraction. The evolution process based on topic intensity reflects the evolution trend of China's tourism poverty alleviation. This study provides a theoretical foundations and technical support for the sustainable development of China's tourism poverty alleviation.
The brain systems of episodic memory and oculomotor control are tightly linked, suggesting a crucial role of eye movements in memory. But little is known about the neural mechanisms of memory formation across eye movements in unrestricted viewing behavior. Here, we leverage simultaneous eye tracking and EEG recording to examine episodic memory formation in free viewing. Participants memorized multi-element events while their EEG and eye movements were concurrently recorded. Each event comprised elements from three categories (face, object, place), with two exemplars from each category, in different locations on the screen. A subsequent associative memory test assessed participants’ memory for the between-category associations that specified each event. We used a deconvolution approach to overcome the problem of overlapping EEG responses to sequential saccades in free viewing. Brain activity was time-locked to the fixation onsets, and we examined EEG power in the theta and alpha frequency bands, the putative oscillatory correlates of episodic encoding mechanisms. Three modulations of fixation-related EEG predicted high subsequent memory performance: 1) theta increase at fixations after between-category gaze transitions, 2) theta and alpha increase at fixations after within-element gaze transitions, 3) alpha decrease at fixations after between-exemplar gaze transitions. Thus, event encoding with unrestricted viewing behavior was characterized by three neural mechanisms, manifested in fixation-locked theta and alpha EEG activity that rapidly turned on and off during the unfolding eye movement sequences. These three distinct neural mechanisms may be the essential building blocks that subserve the buildup of coherent episodic memories during unrestricted viewing behavior.
Background
The number of accidents due to distracted pedestrian is on the rise and many governments and institutions are enacting public policies which restrict texting while walking. However, pedestrians do more than just texting when they use their mobile devices on the go.
Objective
Exploring pedestrian multitasking, this paper aims to examine the effects of mobile device task type on pedestrian performance outcomes.
Method
We performed two studies in lab simulations where 78 participants were asked to perform different tasks on a mobile device (playing a game, reading, writing an email, texting one person, group texting) while performing a pedestrian visual discrimination task while either standing or walking on a treadmill. Behavioral performance as well as neurophysiological data are collected.
Results
Results show that compared to a no-phone control, multitasking with any of the tasks on a mobile device leads to poor performance on a pedestrian visual discrimination task. Playing a game is the most cognitively demanding task and leads to the greatest performance degradation.
Conclusion
Our studies show that multitasking with a mobile device has the potential to negatively impact pedestrian safety, regardless of task type. However, the impacts of different mobile device tasks are not all equivalent. More research is needed to tease out the different effects of these various tasks and to design mobile applications which effectively and safely capture pedestrians’ attention.
Application
Public policy, infrastructure, and smart technologies can be used to mitigate the negative effects of mobile multitasking. A more thorough understanding of mobile device task-specific factors at play can help tailor these counter-measures to better aid distracted pedestrians.
Neural entrainment to musical rhythm is thought to underlie the perception and production of music. In aging populations, the strength of neural entrainment to rhythm has been found to be attenuated, particularly during attentive listening to auditory streams. However, previous studies on neural entrainment to rhythm and aging have often employed artificial auditory rhythms or limited pieces of recorded, naturalistic music, failing to account for the diversity of rhythmic structures found in natural music. As part of larger project assessing a novel music-based intervention for healthy aging, we investigated neural entrainment to musical rhythms in the electroencephalogram (EEG) while participants listened to self-selected musical recordings across a sample of younger and older adults. We specifically measured neural entrainment to the level of musical pulse - quantified here as the phase-locking value (PLV) - after normalizing the PLVs to detected pulse frequencies of each recording. As predicted, we observed strong neural phase-locking to musical pulse, and to the sub-harmonic and harmonic levels of musical meter. Overall, PLVs were not significantly different between older and younger adults. This preserved neural entrainment to musical pulse and rhythm could support the design of music-based interventions that aim to modulate endogenous brain activity via self-selected music for healthy cognitive aging.
Background:
Somatic symptoms, which are common in major depressive disorder (MDD), are associated with a worse prognosis and increased health costs.
Aims:
This randomized, double-blind, placebo-controlled study evaluated the efficacy of aripiprazole augmentation in MDD patients with somatic symptoms.
Methods:
In all, 41 MDD patients with somatic symptoms completed the study. Participants who had been on a stable dose of antidepressants for at least 1 month were randomly assigned to receive an 8-week adjunctive treatment with either aripiprazole or placebo. The initiation dose of aripiprazole was 2 mg/day, which was later adjusted to 1-10 mg/day. The primary endpoint was the change in the Symptom Checklist-90-Revised-Somatization (SCL-90-R-SOM) score. We collected quantitative electroencephalogram data and performed spectral analyses to obtain the absolute power of frequency bands.
Results/outcomes:
The aripiprazole group (n = 20; 2.98 ± 1.75 mg/day) showed a significant improvement in SCL-90-R-SOM scores compared to the placebo group (n = 21; F = 8.56, p = 0.006), without significant differences in changes in depression and anxiety symptoms. Compared to the control, the aripiprazole group showed a greater decrease in total alpha power (F = 7.03, p = 0.01). Changes in frontal alpha power were positively correlated with changes in SCL-90-R-SOM scores in the aripiprazole group (r = 0.53, p = 0.014).
Conclusions/interpretation:
Aripiprazole adjunctive to antidepressants in patients with MDD and somatic symptoms improved somatic symptom severity without significant safety concerns, and this improvement correlated with a decrease in total and frontal alpha power.Trial Registration: https://cris.nih.go.kr; identifier: KCT0004607.
Electroencephalography(EEG), helps to analyze the neuronal activity of a human brain in the form of electrical signals with high temporal resolution in the millisecond range. In the case of e-health cares, the internet is the medium for remote patients.To extract clean clinical information from EEG signals, it is essential to remove unwanted signals, called as artifacts, that are due to different causes including at the time of acquisition. In this piece of work, the authors considered the EEG signal contaminated with ocular artifacts. The clean EEG as well as artifactual EEG are taken from the openly available Mendeley database for verifying the performance of the proposed algorithm.Being the artifactual signal is non-linear and non-stationary the Deep Wavelet Vector Functional Link Network(WVFLN) model is used in this case. The Machine Learning approach has taken a leading role in every field of current research and WVFLN is one of them. For the proof of adaptive nature, the model is designed with EEG as a reference and artifactual EEG as input.To vary the weight and reduce the error, an exponentially weighted Recursive Least Square(RLS) algorithm is used to update the output layer weight of the novel WVFLN model. The Wavelet coefficients are considered in this model with a radial basis function to satisfy the required signal experimentation. It is found that the result is excellent in terms of Mean Square Error(MSE), Root Mean Square Error(RMSE).The MSE and RMSE are found as 0.18μV² and 0.4242μV respectively.Also, the proposed method is compared with the earlier methods to show its efficacy.
Artifact Subspace Reconstruction (ASR) is an adaptive method for the online or offline correction of artifacts comprising multichannel electroencephalography (EEG) recordings. It repeatedly computes a principal component analysis (PCA) on covariance matrices to detect artifacts based on their statistical properties in the component subspace. We adapted the existing ASR implementation by using Riemannian geometry for covariance matrix processing. EEG data that were recorded on smartphone in both outdoors and indoors conditions were used for evaluation (N = 27). A direct comparison between the original ASR and Riemannian ASR (rASR) was conducted for three performance measures: reduction of eye-blinks (sensitivity), improvement of visual-evoked potentials (VEPs) (specificity), and computation time (efficiency). Compared to ASR, our rASR algorithm performed favorably on all three measures. We conclude that rASR is suitable for the offline and online correction of multichannel EEG data acquired in laboratory and in field conditions.
One of the greatest challenges that hinder the decoding and application of electroencephalography (EEG) is that EEG recordings almost always contain artifacts-non-brain signals. Among existing automatic artifact-removal methods, artifact subspace reconstruction (ASR) is an online and real-time capable, component-based method that can effectively remove transient or large-amplitude artifacts. However, the effectiveness of ASR and the optimal choice of its parameter have not been evaluated and reported, especially on real EEG data. This study systematically validates ASR on ten EEG recordings in a simulated driving experiment. Independent component analysis (ICA) is applied to separate artifacts from brain signals to allow a quantitative assessment of ASR's effectiveness in removing various types of artifacts and preserving brain activities. Empirical results show that the optimal ASR parameter is between 10 and 100, which is small enough to remove activities from artifacts and eye-related components and large enough to retain signals from brain-related components. With the appropriate choice of the parameter, ASR can be a powerful and automatic artifact removal approach for offline data analysis or online real-time EEG applications such as clinical monitoring and brain-computer interfaces.
Non-brain contributions to electroencephalo-graphic (EEG) signals, often referred to as artifacts, can hamper the analysis of scalp EEG recordings. This is especially true when artifacts have large amplitudes (e.g., movement artifacts), or occur continuously (like eye-movement artifacts). Offline automated pipelines can detect and reduce artifact in EEG data, but no good solution exists for online processing of EEG data in near real time. Here, we propose the combined use of online artifact subspace reconstruction (ASR) to remove large amplitude transients, and online recursive independent component analysis (ORICA) combined with an independent component (IC) classifier to compute, classify, and remove artifact ICs. We demonstrate the efficacy of the proposed pipeline using 2 EEG recordings containing series of (1) movement and muscle artifacts, and (2) cued blinks and saccades. This pipeline is freely available in the Real-time EEG Source-mapping Toolbox (REST) for MATLAB (The Mathworks, Inc.).
Electroenchephalography (EEG) recordings collected with developmental populations present particular challenges from a data processing perspective. These EEGs have a high degree of artifact contamination and often short recording lengths. As both sample sizes and EEG channel densities increase, traditional processing approaches like manual data rejection are becoming unsustainable. Moreover, such subjective approaches preclude standardized metrics of data quality, despite the heightened importance of such measures for EEGs with high rates of initial artifact contamination. There is presently a paucity of automated resources for processing these EEG data and no consistent reporting of data quality measures. To address these challenges, we propose the Harvard Automated Processing Pipeline for EEG (HAPPE) as a standardized, automated pipeline compatible with EEG recordings of variable lengths and artifact contamination levels, including high-artifact and short EEG recordings from young children or those with neurodevelopmental disorders. HAPPE processes event-related and resting-state EEG data from raw files through a series of filtering, artifact rejection, and re-referencing steps to processed EEG suitable for time-frequency-domain analyses. HAPPE also includes a post-processing report of data quality metrics to facilitate the evaluation and reporting of data quality in a standardized manner. Here, we describe each processing step in HAPPE, perform an example analysis with EEG files we have made freely available, and show that HAPPE outperforms seven alternative, widely-used processing approaches. HAPPE removes more artifact than all alternative approaches while simultaneously preserving greater or equivalent amounts of EEG signal in almost all instances. We also provide distributions of HAPPE's data quality metrics in an 867 file dataset as a reference distribution and in support of HAPPE's performance across EEG data with variable artifact contamination and recording lengths. HAPPE software is freely available under the terms of the GNU General Public License at https://github.com/lcnhappe/happe.
Electroencephalography (EEG) recordings are often contaminated by muscle artifacts. In the literature, a number of methods have been proposed to deal with this problem. Yet most denoising muscle artifact methods are designed for either single-channel EEG or hospital-based, high-density multichannel recordings, not the few-channel scenario seen in most ambulatory EEG instruments. In this paper, we propose utilizing interchannel dependence information seen in the few-channel situation by combining multivariate empirical mode decomposition and canonical correlation analysis (MEMD-CCA). The proposed method, called MEMD-CCA, first utilizes MEMD to jointly decompose the few-channel EEG recordings into multivariate intrinsic mode functions (IMFs). Then, CCA is applied to further decompose the reorganized multivariate IMFs into the underlying sources. Reconstructing the data using only artifact-free sources leads to artifact-attenuated EEG. We evaluated the performance of the proposed method through simulated, semisimulated, and real data. The results demonstrate that the proposed method is a promising tool for muscle artifact removal in the few-channel setting.
In lower mammals, locomotion seems to be mainly regulated by subcortical and spinal networks. On the contrary, recent evidence suggests that in humans the motor cortex is also significantly engaged during complex locomotion tasks. However, a detailed understanding of cortical contribution to locomotion is still lacking especially during stereotyped activities. Here, we show that cortical motor areas finely control leg muscle activation during treadmill stereotyped walking. Using a novel technique based on a combination of Reliable Independent Component Analysis, source localization and effective connectivity, and by combining electroencephalographic (EEG) and electromyographic (EMG) recordings in able-bodied adults we were able to examine for the first time cortical activation patterns and cortico-muscular connectivity including information flow direction. Results not only provided evidence of cortical activity associated with locomotion, but demonstrated significant causal unidirectional drive from contralateral motor cortex to muscles in the swing leg. These insights overturn the traditional view that human cortex has a limited role in the control of stereotyped locomotion, and suggest useful hypotheses concerning mechanisms underlying gait under other conditions.
One sentence summary:
Motor cortex proactively drives contralateral swing leg muscles during treadmill walking, counter to the traditional view of stereotyped human locomotion.
Objective:
Biological and non-biological artifacts cause severe problems when dealing with electroencephalogram (EEG) recordings. Independent component analysis (ICA) is a widely used method for eliminating various artifacts from recordings. However, evaluating and classifying the calculated independent components (IC) as artifact or EEG is not fully automated at present.
Approach:
In this study, we propose a new approach for automated artifact elimination, which applies machine learning algorithms to ICA-based features.
Main results:
We compared the performance of our classifiers with the visual classification results given by experts. The best result with an accuracy rate of 95% was achieved using features obtained by range filtering of the topoplots and IC power spectra combined with an artificial neural network.
Significance:
Compared with the existing automated solutions, our proposed method is not limited to specific types of artifacts, electrode configurations, or number of EEG channels. The main advantages of the proposed method is that it provides an automatic, reliable, real-time capable, and practical tool, which avoids the need for the time-consuming manual selection of ICs during artifact removal.
Neurofeedback is a kind of biofeedback, which teaches self-control of brain functions to subjects by measuring brain waves and providing a feedback signal. Neurofeedback usually provides the audio and or video feedback. Positive or negative feedback is produced for desirable or undesirable brain activities, respectively. In this review, we provided clinical and technical information about the following issues: (1) Various neurofeedback treatment protocols i.e. alpha, beta, alpha/theta, delta, gamma, and theta; (2) Different EEG electrode placements i.e. standard recording channels in the frontal, temporal, central, and occipital lobes; (3) Electrode montages (unipolar, bipolar); (4) Types of neurofeedback i.e. frequency, power, slow cortical potential, functional magnetic resonance imaging, and so on; (5) Clinical applications of neurofeedback i.e. treatment of attention deficit hyperactivity disorder, anxiety, depression, epilepsy, insomnia, drug addiction, schizophrenia, learning disabilities, dyslexia and dyscalculia, autistic spectrum disorders and so on as well as other applications such as pain management, and the improvement of musical and athletic performance; and (6) Neurofeedback softwares. To date, many studies have been conducted on the neurofeedback therapy and its effectiveness on the treatment of many diseases. Neurofeedback, like other treatments, has its own pros and cons. Although it is a non-invasive procedure, its validity has been questioned in terms of conclusive scientific evidence. For example, it is expensive, time-consuming and its benefits are not long-lasting. Also, it might take months to show the desired improvements. Nevertheless, neurofeedback is known as a complementary and alternative treatment of many brain dysfunctions. However, current research does not support conclusive results about its efficacy.
The Electroencephalogram (EEG) is a noninvasive functional brain activity recording method that shows promise for becoming a 3-D cortical imaging modality with high temporal resolution. Currently, most of the tools developed for EEG analysis focus mainly on offline processing. This study introduces and demonstrates the Real-time EEG Source-mapping Toolbox (REST), an extension to the widely distributed EEGLAB environment. REST allows blind source separation of EEG data in real-time using Online Recursive Independent Component Analysis (ORICA), plus near real-time localization of separated sources. Two source localization methods are available to fit equivalent current dipoles or estimate spatial source distributions of selected sources. Selected measures of raw EEG data or component activations (e.g. time series of the data, spectral changes over time, equivalent current dipoles, etc.) can be visualized in near real-time. Finally, this study demonstrates the accuracy and functionality of REST with data from two experiments and discusses some relevant applications.
Background:
Muscle artifacts and electrode noise are an obstacle to interpretation of EEG and other electrophysiological signals. They are often channel-specific and do not fully benefit from component analysis techniques such as ICA, and their presence reduces the dimensionality needed by those techniques. Their high-frequency content may mask or masquerade as gamma band cortical activity.
New method:
The Sparse Time Artifact Removal (STAR) algorithm removes artifacts that are sparse in space and time. The time axis is partitioned into an artifact-free and an artifact-contaminated part, and the correlation structure of the data is estimated from the covariance matrix of the artifact-free part. Artifacts are then corrected by projection of each channel onto the subspace spanned by the other channels.
Results:
The method is evaluated with both simulated and real data, and found to be highly effective in removing or attenuating typical channel-specific artifacts.
Comparison with existing methods:
In contrast to the widespread practice of trial removal or channel removal or interpolation, very few data are lost. In contrast to ICA or other linear techniques, processing is local in time and affects only the artifact part, so most of the data are identical to the unprocessed data and the full dimensionality of the data is preserved.
Conclusions:
STAR complements other linear component analysis techniques, and can enhance their ability to discover weak sources of interest by increasing the number of effective noise-free channels.
Goal:
We present and evaluate a wearable high-density dry-electrode EEG system and an open-source software framework for online neuroimaging and state classification.
Methods:
The system integrates a 64-channel dry EEG form factor with wireless data streaming for online analysis. A real-time software framework is applied, including adaptive artifact rejection, cortical source localization, multivariate effective connectivity inference, data visualization, and cognitive state classification from connectivity features using a constrained logistic regression approach (ProxConn). We evaluate the system identification methods on simulated 64-channel EEG data. Then, we evaluate system performance, using ProxConn and a benchmark ERP method, in classifying response errors in nine subjects using the dry EEG system.
Results:
Simulations yielded high accuracy (AUC = 0.97 ± 0.021) for real-time cortical connectivity estimation. Response error classification using cortical effective connectivity [short-time direct-directed transfer function (sdDTF)] was significantly above chance with similar performance (AUC) for cLORETA (0.74 ±0.09) and LCMV (0.72 ±0.08) source localization. Cortical ERP-based classification was equivalent to ProxConn for cLORETA (0.74 ±0.16) but significantly better for LCMV (0.82 ±0.12) .
Conclusion:
We demonstrated the feasibility for real-time cortical connectivity analysis and cognitive state classification from high-density wearable dry EEG.
Significance:
This paper is the first validated application of these methods to 64-channel dry EEG. This study addresses a need for robust real-time measurement and interpretation of complex brain activity in the dynamic environment of the wearable setting. Such advances can have broad impact in research, medicine, and brain-computer interfaces. The pipelines are made freely available in the open-source SIFT and BCILAB toolboxes.
Electroencephalogram (EEG) is susceptible to various nonneural physiological artifacts. Automatic artifact removal from EEG data remains a key challenge for extracting relevant information from brain activities. To adapt to variable subjects and EEG acquisition environments, this paper presents an automatic online artifact removal method based on a priori artifact information. The combination of discrete wavelet transform and independent component analysis (ICA), wavelet-ICA, was utilized to separate artifact components. The artifact components were then automatically identified using a priori artifact information, which was acquired in advance. Subsequently, signal reconstruction without artifact components was performed to obtain artifact-free signals. The results showed that, using this automatic online artifact removal method, there were statistical significant improvements of the classification accuracies in both two experiments, namely, motor imagery and emotion recognition.
This paper presents an extensive review on the artifact removal algorithms used to remove the main sources of interference encountered in the electroencephalogram (EEG), specifically ocular, muscular and cardiac artifacts. We first introduce background knowledge on the characteristics of EEG activity, of the artifacts and of the EEG measurement model. Then, we present algorithms commonly employed in the literature and describe their key features. Lastly, principally on the basis of the results provided by various researchers, but also supported by our own experience, we compare the state-of-the-art methods in terms of reported performance, and provide guidelines on how to choose a suitable artifact removal algorithm for a given scenario. With this review we have concluded that, without prior knowledge of the recorded EEG signal or the contaminants, the safest approach is to correct the measured EEG using independent component analysis-to be precise, an algorithm based on second-order statistics such as second-order blind identification (SOBI). Other effective alternatives include extended information maximization (InfoMax) and an adaptive mixture of independent component analyzers (AMICA), based on higher order statistics. All of these algorithms have proved particularly effective with simulations and, more importantly, with data collected in controlled recording conditions. Moreover, whenever prior knowledge is available, then a constrained form of the chosen method should be used in order to incorporate such additional information. Finally, since which algorithm is the best performing is highly dependent on the type of the EEG signal, the artifacts and the signal to contaminant ratio, we believe that the optimal method for removing artifacts from the EEG consists in combining more than one algorithm to correct the signal using multiple processing stages, even though this is an option largely unexplored by researchers in the area.
Low frequency signals recorded from non-invasive electroencephalography (EEG), in particular movement-related cortical potentials (MRPs), are associated with preparation and execution of movement and thus present a target for use in brain-machine interfaces. We investigated the ability to decode movement intent from delta-band (0.1-4 Hz) EEG recorded immediately before movement execution in healthy volunteers. We used data from epochs starting 1.5 s before movement onset to classify future movements into one of three classes: stand-up, sit-down, or quiet. We assessed classification accuracy in both externally triggered and self-paced paradigms. Movement onset was determined from electromyography (EMG) recordings synchronized with EEG signals. We employed an artifact subspace reconstruction (ASR) algorithm to eliminate high amplitude noise before building our time-embedded EEG features. We applied local Fisher's discriminant analysis to reduce the dimensionality of our spatio-temporal features and subsequently used a Gaussian mixture model classifier for our three class problem. Our results demonstrate significantly better than chance classification accuracy (chance level = 33.3%) for the self-initiated (78.0 ± 2.6%) and triggered (74.7 ± 5.7%) paradigms. Surprisingly, we found no significant difference in classification accuracy between the self-paced and cued paradigms when using the full set of non-peripheral electrodes. However, accuracy was significantly increased for self-paced movements when only electrodes over the primary motor area were used. Overall, this study demonstrates that delta-band EEG recorded immediately before movement carries discriminative information regarding movement type. Our results suggest that EEG-based classifiers could improve lower-limb neuroprostheses and neurorehabilitation techniques by providing earlier detection of movement intent, which could be used in robot-assisted strategies for motor training and recovery of function.
Objective: In a systematic study of gamma activity in neuro-psychiatric disease, we unexpectedly observed distinctive, apparently persistent, electroencephalogram (EEG) spectral peaks in the gamma range (25–100 Hz). Our objective, therefore, was to examine the incidence, distribution and some of the characteristics of these peaks.
Methods: High sample-rate, 128-channel, EEG was recorded in 603 volunteers (510 with neuropsychiatric disorders, 93 controls), whilst performing cognitive tasks, and converted to power spectra. Peaks of spectral power, including in the gamma range, were determined algorithmically for all electrodes. To determine if peaks were stable, 24-h ambulatory recordings were obtained from 16 subjects with peaks. In 10 subjects, steady-state responses to stimuli at peak frequency were compared with off-peak-frequency stimulation to determine if peaks were a feature of underlying network resonances and peaks were evaluated with easy and hard versions of oddball tasks to determine if peaks might be influenced by mental effort.
Results: 57% of 603 subjects exhibited peaks >2 dB above trough power at or above 25 Hz. Larger peaks (>5 dB) were present in 13% of subjects. Peaks were distributed widely over the scalp, more frequent centrally. Peaks were present through the day and were suppressed by slow-wave-sleep. Steady-state responses were the same with on- or off-peak sensory stimulation. In contrast, mental effort resulted in reductions in power and frequency of gamma peaks, although the suppression did not correlate with level of effort.
Conclusions: Gamma EEG can be expressed constitutively as concentrations of power in narrow or wide frequency bands that play an, as yet, unknown role in cognitive activity.
Significance: These findings expand the described range of rhythmic EEG phenomena. In particular, in addition to evoked, induced and sustained gamma band activity, gamma activity can be present constitutively in spectral peaks.
Objective:
EEG artifacts of non-neural origin can be separated from neural signals by independent component analysis (ICA). It is unclear (1) how robustly recently proposed artifact classifiers transfer to novel users, novel paradigms or changed electrode setups, and (2) how artifact cleaning by a machine learning classifier impacts the performance of brain-computer interfaces (BCIs).
Approach:
Addressing (1), the robustness of different strategies with respect to the transfer between paradigms and electrode setups of a recently proposed classifier is investigated on offline data from 35 users and 3 EEG paradigms, which contain 6303 expert-labeled components from two ICA and preprocessing variants. Addressing (2), the effect of artifact removal on single-trial BCI classification is estimated on BCI trials from 101 users and 3 paradigms.
Main results:
We show that (1) the proposed artifact classifier generalizes to completely different EEG paradigms. To obtain similar results under massively reduced electrode setups, a proposed novel strategy improves artifact classification. Addressing (2), ICA artifact cleaning has little influence on average BCI performance when analyzed by state-of-the-art BCI methods. When slow motor-related features are exploited, performance varies strongly between individuals, as artifacts may obstruct relevant neural activity or are inadvertently used for BCI control.
Significance:
Robustness of the proposed strategies can be reproduced by EEG practitioners as the method is made available as an EEGLAB plug-in.
Independent component analysis (ICA) can find distinct sources of electroencephalographic (EEG) activity, both brain-based and artifactual, and has become a common pre-preprocessing step in analysis of EEG data. Distinction between brain and non-brain independent components (ICs) accounting for, e.g., eye or muscle activities is an important step in the analysis. Here we present a fully automated method to identify eye-movement related EEG components by analyzing the spatial distribution of their scalp projections (scalp maps). The EyeCatch method compares each input scalp map to a database of eye-related IC scalp maps obtained by data-mining over half a million IC scalp maps obtained from 80,006 EEG datasets associated with a diverse set of EEG studies and paradigms. To our knowledge this is the largest sample of IC scalp maps that has ever been analyzed. Our result show comparable performance to a previous state-of-art semi-automated method, CORRMAP, while eliminating the need for human intervention.
Objective:
The past two decades have seen dramatic progress in our ability to model brain signals recorded by electroencephalography, functional near-infrared spectroscopy, etc., and to derive real-time estimates of user cognitive state, response, or intent for a variety of purposes: to restore communication by the severely disabled, to effect brain-actuated control and, more recently, to augment human-computer interaction. Continuing these advances, largely achieved through increases in computational power and methods, requires software tools to streamline the creation, testing, evaluation and deployment of new data analysis methods.
Approach:
Here we present BCILAB, an open-source MATLAB-based toolbox built to address the need for the development and testing of brain-computer interface (BCI) methods by providing an organized collection of over 100 pre-implemented methods and method variants, an easily extensible framework for the rapid prototyping of new methods, and a highly automated framework for systematic testing and evaluation of new implementations.
Main results:
To validate and illustrate the use of the framework, we present two sample analyses of publicly available data sets from recent BCI competitions and from a rapid serial visual presentation task. We demonstrate the straightforward use of BCILAB to obtain results compatible with the current BCI literature.
Significance:
The aim of the BCILAB toolbox is to provide the BCI community a powerful toolkit for methods research and evaluation, thereby helping to accelerate the pace of innovation in the field, while complementing the existing spectrum of tools for real-time BCI experimentation, deployment and use.
Independent component analysis (ICA) and blind source separation (BSS) methods are increasingly used to separate individual brain and non-brain source signals mixed by volume conduction in electroencephalographic (EEG) and other electrophysiological recordings. We compared results of decomposing thirteen 71-channel human scalp EEG datasets by 22 ICA and BSS algorithms, assessing the pairwise mutual information (PMI) in scalp channel pairs, the remaining PMI in component pairs, the overall mutual information reduction (MIR) effected by each decomposition, and decomposition 'dipolarity' defined as the number of component scalp maps matching the projection of a single equivalent dipole with less than a given residual variance. The least well-performing algorithm was principal component analysis (PCA); best performing were AMICA and other likelihood/mutual information based ICA methods. Though these and other commonly-used decomposition methods returned many similar components, across 18 ICA/BSS algorithms mean dipolarity varied linearly with both MIR and with PMI remaining between the resulting component time courses, a result compatible with an interpretation of many maximally independent EEG components as being volume-conducted projections of partially-synchronous local cortical field activity within single compact cortical domains. To encourage further method comparisons, the data and software used to prepare the results have been made available (http://sccn.ucsd.edu/wiki/BSSComparison).
Electroencephalographic (EEG) correlates of driving performance were studied using an event-related lane-departure paradigm.
High-density EEG data were analyzed using independent component analysis (ICA) and Fourier analysis. Across subjects and sessions,
when reaction time to lane-departure events increased, several clusters of independent component activities in the occipital,
posterior parietal, and middle temporal cortex showed tonic power increases in the delta, theta, and alpha bands. The strongest
of these tonic power increases occurred in the alpha band in the occipital and parietal regions. Other independent component
clusters in the somatomotor and frontal regions showed less or no significant increase in all frequency bands as RT increased.
This study demonstrates additional evidence of the close and specific links between cortical brain activities (via changes
in EEG spectral power) and performance (reaction time) during sustained-attention tasks. These results may also provide insights
into the development of human-computer interfaces for countermeasures for drowsy driving.
An extension of the infomax algorithm of Bell and Sejnowski (1995) is presented that is able blindly to separate mixed signals with sub- and supergaussian source distributions. This was achieved by using a simple type of learning rule first derived by Girolami (1997) by choosing negentropy as a projection pursuit index. Parameterized probability distributions that have sub- and supergaussian regimes were used to derive a general learning rule that preserves the simple architecture proposed by Bell and Sejnowski (1995), is optimized using the natural gradient by Amari (1998), and uses the stability analysis of Cardoso and Laheld (1996) to switch between sub- and supergaussian regimes. We demonstrate that the extended infomax algorithm is able to separate 20 sources with a variety of source distributions easily. Applied to high-dimensional data from electroencephalographic recordings, it is effective at separating artifacts such as eye blinks and line noise from weaker electrical signals that arise from sources in the brain.
