Article
To read the full-text of this research, you can request a copy directly from the authors.

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.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Later, an evaluation study on 10 EEG files revealed that the previously suggested values were too aggressive, and a new optimal range between 10 and 100 was recommended [10]. As this range was still broad, the same researchers later provided an optimal value for k lying between 20 and 30 by extending their analysis to additional 10 EEG files [17]. A key thing to note here is that the assessment was based on the Independent Components (IC) extracted by ICA with and without ASR processing. ...
... A Signalto-Noise ratio (SNR, defined in [8]), as the quality metric derived on the Steady-State Visually Evoked Potentials EEG data, is used to identify the optimal k. Like the previous study [17], the considered quality metric depends on several other parameters (e.g., window size for Fourier analysis, number of samples etc.,) and not just on the ASR parameter k; therefore, more direct analysis is currently missing in the literature. Moreover, the number of EEG files considered in both of these studies are 20 and 6, respectively, and such smaller sample sizes might not be helpful to derive a statistically robust conclusion. ...
... The performance of ASR cleaning is strongly dependent on the user-defined tunable ASR parameter k. While the previous studies evaluated this parameter using metrics (Correlation coefficient of ICs [10,17], or SNR [14]), a direct validation against the labelled ground truth was still missing. Moreover, the results concluded from previous works might not be considered statistically significant due to the smaller set of EEG files (10, 20 and 6, respectively). ...
Conference Paper
Artifacts preprocessing in EEG is remarkably significant to extract reliable neural responses in the downstream analysis. A recently emerging powerful preprocessing tool among the EEG community is Artifacts Subspace Reconstruction (ASR). ASR is an unsupervised machine learning algorithm to identify and correct the transient-like non-stationary noisy samples. ASR is fully automatic, therefore, suitable for online applications. However, the performance of ASR is strongly dependent on the user-defined hyperparameter k. A poor choice of k might lead to severe performance degradation. In this work, we benchmark the performance of ASR against its parameter k. Toward this goal, we used the Temple University Hospital EEG Artifact Corpus (TUAR), which consists of 310 EEG files recorded in clinical settings from epileptic patients. Remarkably, these files are annotated for artifacts by trained personnel with a high inter-rater agreement score (κ> 0.8). Considering these reliable labels as ground truth, ASR has shown the best performance in artifacts cleaning with k ranging between 20 and 40.
... While ASR requires no training to identify such artifacts, its performance is highly dependent on the choice of its user-defined parameters. The most crucial one, ASR Rejection Threshold Parameter k, was extensively studied in [7], where a value between 20 and 30 is suggested for high-density EEG data. Moreover, as a co-variance-based component method, ASR cannot be applied to single channel EEG. ...
... Moreover, as a co-variance-based component method, ASR cannot be applied to single channel EEG. For an effective artifact removal, a standard EEG system with at least 20 channels is recommended [7]. Therefore, it is not yet clear whether ASR could efficiently remove artifacts on data from low-density wearable EEG systems, and if so what could be its optimal parameters. ...
... We then varied ASR parameters to find the optimal ones for artifact removal. Differently from previous studies [7], not only we systematically varied ASR key parameter k, but we also compared the two main processing modes of ASR: removing artifacts from bad segments (Correction) or rejecting bad segments (Removal). We tested the ASR performance by comparing SSVEPs of ASR-cleaned data from artifact-free and artifacted sessions at all stimulation frequencies. ...
Conference Paper
Lightweight , minimally-obtrusive mobile EEG systems with a small number of electrodes (i.e., low-density) allow for convenient monitoring of the brain activity in out-of-the-lab conditions. However, they pose a higher risk for signal contamination with non-stereotypical artifacts due to hardware limitations and the challenging environment where signals are collected. A promising solution is Artifacts Subspace Reconstruction (ASR), a component-based approach that can automatically remove non-stationary transient-like artifacts in EEG data. Since ASR has only been validated with high-density systems, it is unclear whether it is equally efficient on low-density portable EEG. This paper presents a complete analysis of ASR performance based on clean and contaminated datasets acquired with BioWolf, an Ultra-Low-Power system featuring only eight channels, during SSVEP sessions recorded from six adults. Empirical results show that even with such few channels, ASR efficiently corrects artifacts, enabling an overall enhancement of up to 40% in SSVEP response. Furthermore, by choosing the optimal ASR parameters on a single-subject basis, SSVEP response can be further increased to more than 45%. These results suggest that ASR is a viable and robust method for online automatic artifact correction with low-density BCI systems in real-life scenarios.
... Thanks to its efficient artifact removal, ASR is now considered as one of the default preprocessing algorithms within the EEGLAB framework. However, ASR has only been evaluated on adult EEG data thus far (Blum et al., 2019;Chang et al., 2020;Mullen et al., 2015). For the first time, in this work, we evaluate ASR on noisier developmental EEG data and propose it as one of the core blocks in our pipeline. ...
... In addition, we propose a calibration procedure for adapting ASR algorithm to developmental data. ASR processes artifacts in three steps that are briefly described as follows (for more detailed technical documentation, please refer to Jung, 2016, andChang et al., 2020). ...
... It can be observed that a lower k implies a lower threshold and therefore a strict artifact detection (i.e. more artifacts are detected); a higher k implies a looser cleaning of the data (i.e. less artifacts are detected). For adult EEG, the optimal k values lie in the range between 20 and 30 (Chang et al., 2020). As mentioned before, to the best of our knowledge, the ASR parameter k has never been evaluated on developmental data. ...
Article
Full-text available
Electroencephalography (EEG) is arising as a valuable method to investigate neurocognitive functions shortly after birth. However, obtaining high-quality EEG data from human newborn recordings is challenging. Compared to adults and older infants, datasets are typically much shorter due to newborns’ limited attentional span and much noisier due to non-stereotyped artifacts mainly caused by uncontrollable movements. We propose Newborn EEG Artifact Removal (NEAR), a pipeline for EEG artifact removal designed explicitly for human newborns. NEAR is based on two key steps: 1) A novel bad channel detection tool based on the Local Outlier Factor (LOF), a robust outlier detection algorithm; 2) A parameter calibration procedure for adapting to newborn EEG data the algorithm Artifacts Subspace Reconstruction (ASR), developed for artifact removal in mobile adult EEG. Tests on simulated data showed that NEAR outperforms existing methods in removing representative newborn non-stereotypical artifacts. NEAR was validated on two pediatric populations (newborns and 9-month-old infants) recorded with two different experimental designs (frequency-tagging and ERP). Results show that NEAR artifact removal successfully reproduces established EEG responses from noisy datasets, with a higher statistical significance than the one obtained by existing artifact removal methods. The EEGLAB-based NEAR pipeline is freely available at https://github.com/vpKumaravel/NEAR.
... Generally, it is recommended to use 64 channels or more, and in movement studies, it is good to increase the number of channels with increasing movement intensity [5] to provide higher degrees of freedom for ICA to explain the increasing numbers of potential artifactual sources. Another method found to improve ICA decomposition is first cleaning the data with artifact subspace reconstruction (ASR) [52], which we discuss below. ...
... Another method used in movement studies is ASR [12,62,63]. The ASR method has several advantages including the automated removal of artifactual components, its usability for online applications, and the ability to remove transient or large-amplitude artifacts that the ICA method struggles with [52]. This method is relatively new, and its application to movement EEG data is currently poorly evaluated. ...
... In the end, the ASR method rejects the artifact components in each time 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 window if the principal component is larger than the rejection threshold. Subsequently, the final reconstruction of the cleaned signals from the remaining data was computed [20,52]. Reimannian ASR is an improved version of the ASR method that uses Reimannian methods for covariance matrices computation, which has been shown to be beneficial for artifact removal [64]. ...
Article
Electroencephalography (EEG) is a non-invasive technique used to record cortical neurons' electrical activity using electrodes placed on the scalp. It has become a promising avenue for research beyond state-of-the-art EEG research that is conducted under static conditions. EEG signals are always contaminated by artifacts and other physiological signals. Artifact contamination increases with the intensity of movement. In the last decade (since 2010), researchers have started to implement EEG measurements in dynamic setups to increase the overall ecological validity of the studies. Many different methods are used to remove non-brain activity from the EEG signal, and there are no clear guidelines on which method should be used in dynamic setups and for specific movement intensities. Currently, the most common methods for removing artifacts in movement studies are methods based on independent component analysis (ICA). However, the choice of method for artifact removal depends on the type and intensity of movement, which affects the characteristics of the artifacts and the EEG parameters of interest. When dealing with EEG under non-static conditions, special care must be taken already in the designing period of an experiment. Software and hardware solutions must be combined to achieve sufficient removal of unwanted signals from EEG measurements. We have provided recommendations for the use of each method depending on the intensity of the movement and highlighted the advantages and disadvantages of the methods. However, due to the current gap in the literature, further development and evaluation of methods for artifact removal in EEG data during locomotion is needed.
... Several researchers have proposed various countermeasures to remove signal contamination caused by, for example, eye blinks/movements, muscle activity, and line noise. One commonly adopted countermeasure is blind source separation (BSS) (Fitzgibbon et al., 2007;Joyce et al., 2004;Jung et al., 2000), in which various statistical assumptions, such as independence (Jung et al., 2000;Makeig et al., 1996), correlation (de Cheveigné et al., 2018;Lin et al., 2018), covariance (Chang et al., 2020), or stationarity (Maddirala and Veluvolu, 2021;von Bunau et al., 2010), are used to extract the underlying sources from their mixtures at sensors. Artifact subspace reconstruction (ASR) (Chang et al., 2020), for instance, assumes that the EEG data share the same principal component space, projects raw EEG signals onto this feature space, and then reconstructs signals by preserving those projected principal components whose variance does not exceed a certain predefined threshold. ...
... One commonly adopted countermeasure is blind source separation (BSS) (Fitzgibbon et al., 2007;Joyce et al., 2004;Jung et al., 2000), in which various statistical assumptions, such as independence (Jung et al., 2000;Makeig et al., 1996), correlation (de Cheveigné et al., 2018;Lin et al., 2018), covariance (Chang et al., 2020), or stationarity (Maddirala and Veluvolu, 2021;von Bunau et al., 2010), are used to extract the underlying sources from their mixtures at sensors. Artifact subspace reconstruction (ASR) (Chang et al., 2020), for instance, assumes that the EEG data share the same principal component space, projects raw EEG signals onto this feature space, and then reconstructs signals by preserving those projected principal components whose variance does not exceed a certain predefined threshold. Another well-grounded and effective BSS method is independent component analysis (ICA) (Jung et al., 2000;Makeig et al., 1996), which separates instantaneously and linearly mixed recordings into mutually independent components (ICs). ...
... To verify that the reconstructed signals were brain activities and still possessed the characteristics of EEG signals, after completing artifact removal processing, ICA and ICLabel(Pion-Tonachini et al., 2019) were applied to the reconstructed signals to see if the number of non-brain ICs decreased and whether the number of extractable brain ICs increased.The artifact removal performance of the proposed model was compared to that of three other methods(Fig. 3C): bandpass filter (frequency range = 1-50 Hz), a component-based method (i.e., ASR(Chang et al., 2020) with burst cutoff parameter = 5), and a DL-based method (i.e.,1D- ResCNN (Sun et al., 2020) withEEGdenoiseNet (Zhang et al., 2021) and hyperparameters set to default values). ...
Preprint
Electroencephalography (EEG) signals are often contaminated with artifacts. It is imperative to develop a practical and reliable artifact removal method to prevent misinterpretations of neural signals and underperformance of brain-computer interfaces. This study developed a new artifact removal method, IC-U-Net, which is based on the U-Net architecture for removing pervasive EEG artifacts and reconstructing brain sources. The IC-U-Net was trained using mixtures of brain and non-brain sources decomposed by independent component analysis and employed an ensemble of loss functions to model complex signal fluctuations in EEG recordings. The effectiveness of the proposed method in recovering brain sources and removing various artifacts (e.g., eye blinks/movements, muscle activities, and line/channel noises) was demonstrated in a simulation study and three real-world EEG datasets collected at rest and while driving and walking. IC-U-Net is user-friendly and publicly available, does not require parameter tuning or artifact type designations, and has no limitations on channel numbers. Given the increasing need to image natural brain dynamics in a mobile setting, IC-U-Net offers a promising end-to-end solution for automatically removing artifacts from EEG recordings.
... Generally, it is recommended to use 64 channels or more, and in movement studies, it is good to increase the number of channels with increasing movement intensity [5] to provide higher degrees of freedom for ICA to explain the increasing numbers of potential artifactual sources. Another method found to improve ICA decomposition is first cleaning the data with ASR [52], which we discuss below. ...
... Another method used in movement studies is ASR [12,62,63]. The ASR method has several advantages including the automated removal of artifactual components, its usability for online applications, and the ability to remove transient or large-amplitude artifacts that the ICA method struggles with [52]. This method is relatively new, and its application to movement EEG data is currently poorly evaluated. ...
... In the end, the ASR method rejects the artifact components in each time window if the principal component is larger than the rejection threshold. Subsequently, the final reconstruction of the cleaned signals from the remaining data was computed [20,52]. Reimannian ASR is an improved version of the ASR method that uses Reimannian methods for covariance matrices computation, which has been shown to be beneficial for artifact removal [64]. ...
Article
Full-text available
Electroencephalography (EEG) is a non-invasive technique used to record cortical neurons' electrical activity using electrodes placed on the scalp. It has become a promising avenue for research beyond state-of-the-art EEG research that is conducted under static conditions. EEG signals are always contaminated by artifacts and other physiological signals. Artifact contamination increases with the intensity of movement. In the last decade (since 2010), researchers have started to implement EEG measurements in dynamic setups to increase the overall ecological validity of the studies. Many different methods are used to remove non-brain activity from the EEG signal, and there are no clear guidelines on which method should be used in dynamic setups and for specific movement intensities. Currently, the most common methods for removing artifacts in movement studies are methods based on independent component analysis (ICA). However, the choice of method for artifact removal depends on the type and intensity of movement, which affects the characteristics of the artifacts and the EEG parameters of interest. When dealing with EEG under non-static conditions, special care must be taken already in the designing period of an experiment. Software and hardware solutions must be combined to achieve sufficient removal of unwanted signals from EEG measurements. We have provided recommendations for the use of each method depending on the intensity of the movement and highlighted the advantages and disadvantages of the methods. However, due to the current gap in the literature, further development and evaluation of methods for artifact removal in EEG data during locomotion is needed.
... Using the PrepPipeline plug-in [13], the line noise was removed, noisy channels were identified and interpolated for consistency, and the channels were rereferenced to the average of all electrodes. The artefact subspace reconstruction (ASR) plug-in [14] was used to correct large artefacts and data discontinuities. Finally, wavelet-enhanced independent component analysis (wICA) was used to decompose the EEG signal and remove any large artefacts left based on a threshold [15]. ...
... The Fast-Fourier transform (FFT) was used to calculate the activity separately for each event and baseline from the same trial at each band of interest: delta (2-4 Hz), theta (4-8 Hz), alpha (8)(9)(10)(11)(12)(13)(14), beta (14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), and gamma (30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45). Inverse modelling was done using the exact low-resolution brain electromagnetic tomography (eLORETA) [21] to estimate the values of the sources for each frequency within a band. ...
... The Fast-Fourier transform (FFT) was used to calculate the activity separately for each event and baseline from the same trial at each band of interest: delta (2-4 Hz), theta (4-8 Hz), alpha (8)(9)(10)(11)(12)(13)(14), beta (14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), and gamma (30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45). Inverse modelling was done using the exact low-resolution brain electromagnetic tomography (eLORETA) [21] to estimate the values of the sources for each frequency within a band. ...
Conference Paper
A microsleep (MS) is a complete lapse of responsiveness due to an episode of brief sleep (≲ 15 s) with eyes partially or completely closed. MSs are highly correlated with the risk of car accidents, severe injuries, and death. To investigate EEG changes during MSs, we used a 2D continuous visuomotor tracking (CVT) task and eye-video to identify MSs in 20 subjects performing the 50-min task. Following pre-processing, FFT spectral analysis was used to calculate the activity in the EEG delta, theta, alpha, beta, and gamma bands, followed by eLORETA for source reconstruction. A group statistical analysis was performed to compare the change in activity over EEG bands of an MS to its baseline. After correction for multiple comparisons, we found maximum increases in delta, theta, and alpha activities over the frontal lobe, and beta over the parietal and occipital lobes. There were no significant changes in the gamma band, and no significant decreases in any band. Our results are in agreement with previous studies which reported increased alpha activity in MSs. However, this is the first study to have reported increased beta activity during MSs, which, due to the usual association of beta activity with wakefulness, was unexpected.
... EEG signals represent synchronized electrical pulses from masses of neurons interacting with one other. Brain rhythms are primarily (1 -4 Hz) depicts lowest frequency waves, followed by theta band (4 -7Hz), alpha band (8)(9)(10)(11)(12)(13), beta band (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) and gamma band (30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45). ...
... divided into five frequency bands, differentiated via their morphological and functional aspects. These are majorly classified into five frequency bands: delta (1)(2)(3)(4), theta (4-7 Hz), alpha (8)(9)(10)(11)(12)(13), beta (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), and gamma (30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45). Figure. 2 displays the five brain rhythms. ...
... The subsequent step involves preprocessing and segmentation of EEG signals into epochs of 5 seconds. Then the extracted segments are passed for frequency decomposition into five frequency bands comprising delta (1-4 Hz), theta (4 -7 Hz), alpha (8-13 Hz), beta (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), and gamma (30)(31)(32)(33)(34)(35) bands. This is followed by the construction of brain networks with a threshold that retains the significant connections. ...
Preprint
Full-text available
Neural oscillations are the rich source to understand cognition, perception, and emotions. Decades of research on brain oscillations have primarily discussed neural signatures for the western classification of emotions. Despite this, the Indian ancient treatise on emotions popularly known as Rasas has remained unexplored. In this study, we collected Electroencephalography (EEG) encodings while participants watched nine emotional movie clips corresponding to nine Rasas . The key objective of this study is to identify the brain waves that could distinguish between Rasas . Therefore, we decompose the EEG signals into five primary frequency bands comprising delta (1-4 Hz), theta (4-7 Hz), alpha (8-13 Hz), beta (13-30 Hz), and gamma (30-45 Hz). We construct the functional networks from EEG time-series data and subsequently utilize the fourteen graph-theoretical measures to compute the features. Random Forest models are trained on the extracted features, and we present our findings based on classifier predictions. We observe slow (delta) and fast brain waves (beta and gamma) exhibited the maximum discriminating features between Rasas , whereas alpha and theta bands showed fewer distinguishable pairs. Out of nine Rasas , Sringaram, Bibhatsam, and Bhayanakam displayed the most distinguishing characteristics from other Rasas . Interestingly, our results are consistent with the previous studies, which highlight the significant role of higher frequency oscillations for the classification of emotions. Our finding on the alpha band is consistent with the previous study, which reports the maximum similarity in brain networks across emotions in the alpha band. This research contributes to the pioneering work on Indian Rasas utilizing brain responses.
... Firstly, the low-frequency components were eliminated by applying a Butterworth-type Band Pass digital filter with order 6 of zero phase, and with cutoff frequencies between 0.1 and 30 Hz. Secondly, channels were removed according to the criteria reported in [30]: flat for more than 5 s, maximum acceptable high-frequency noise standard deviation of 4, minimum acceptable correlation with nearby channels of 0.8. Thirdly, Artifact Subspace Reconstruction (ASR) bad burst correction was performed in order to remove bad data periods with transient or largeamplitude artifacts that exceeded 20 times the standard deviation of the calibrated data [30]. ...
... Secondly, channels were removed according to the criteria reported in [30]: flat for more than 5 s, maximum acceptable high-frequency noise standard deviation of 4, minimum acceptable correlation with nearby channels of 0.8. Thirdly, Artifact Subspace Reconstruction (ASR) bad burst correction was performed in order to remove bad data periods with transient or largeamplitude artifacts that exceeded 20 times the standard deviation of the calibrated data [30]. Fourthly, Independent Component Analysis (ICA) was applied with RunICA function. ...
Article
Full-text available
Tinnitus is an auditory condition that causes humans to hear a sound anytime, anywhere. Chronic and refractory tinnitus is caused by an over synchronization of neurons. Sound has been applied as an alternative treatment to resynchronize neuronal activity. To date, various acoustic therapies have been proposed to treat tinnitus. However, the effect is not yet well understood. Therefore, the objective of this study is to establish an objective methodology using electroencephalography (EEG) signals to measure changes in attentional processes in patients with tinnitus treated with auditory discrimination therapy (ADT). To this aim, first, event-related (de-) synchronization (ERD/ERS) responses were mapped to extract the levels of synchronization related to the auditory recognition event. Second, the deep representations of the scalograms were extracted using a previously trained Convolutional Neural Network (CNN) architecture (MobileNet v2). Third, the deep spectrum features corresponding to the study datasets were analyzed to investigate performance in terms of attention and memory changes. The results proved strong evidence of the feasibility of ADT to treat tinnitus, which is possibly due to attentional redirection.
... Artifact subspace reconstruction (ASR; Kothe and Jung, 2016) is a component-based artifact removal method that can be used to clean large-variance signal components based on thresholds compared to clean baseline data and subsequent reconstruction of EEG channel data. By pre-conditioning EEG channel data and removing eye movement and muscle artifacts ahead of ICA, it is possible to improve ICA decomposition quality (Chang et al., 2019). ...
... High pass filtering mobile EEG data provides a partial solution (Winkler et al., 2015), with a 1-2 Hz high pass filter improving subsequent ICA decomposition results, but a number of alternative signal processing solutions have been implemented in mobile EEG studies. Adaptive filtering (Kilicarslan et al., 2016), template regression (Gwin et al., 2010), and component-based statistical decomposition methods, including artifact subspace reconstruction (Chang et al., 2018(Chang et al., , 2019, have been used to eliminate motion artifacts from mobile EEG data at relatively slow gait speeds (<1.0 m/s; Gwin et al., 2010;Wagner et al., 2012Wagner et al., , 2016Bradford et al., 2016Bradford et al., , 2019Oliveira et al., 2017a,b;Bradford et al., 2019), but gait speeds closer to, and in excess of, preferred human walking speed (1.4 m/s; Bohannon, 1997) have remained challenging and have required novel solutions. ...
Article
Full-text available
Walking or running in real-world environments requires dynamic multisensory processing within the brain. Studying supraspinal neural pathways during human locomotion provides opportunities to better understand complex neural circuity that may become compromised due to aging, neurological disorder, or disease. Knowledge gained from studies examining human electrical brain dynamics during gait can also lay foundations for developing locomotor neurotechnologies for rehabilitation or human performance. Technical barriers have largely prohibited neuroimaging during gait, but the portability and precise temporal resolution of non-invasive electroencephalography (EEG) have expanded human neuromotor research into increasingly dynamic tasks. In this narrative mini-review, we provide a (1) brief introduction and overview of modern neuroimaging technologies and then identify considerations for (2) mobile EEG hardware, (3) and data processing, (4) including technical challenges and possible solutions. Finally, we summarize (5) knowledge gained from human locomotor control studies that have used mobile EEG, and (6) discuss future directions for real-world neuroimaging research.
... Electroencephalogram (EEG) is a voltage test recording the electrical activity of the neurons in the brain. EEG is obtained in various frequencies including delta (1)(2)(3)(4), theta (4)(5)(6)(7)(8), alpha , beta (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), and gamma (greater than 30 Hz) [1]. EEG brain rhythms with different frequency ranges are shown in Table 1 [2]. ...
... In 2019, Chang et al. [22] have evaluated artifact subspace reconstruction (ASR) during the simulated driving experiments. The IC classifier and ICA separated the artifacts for assessing the efficiency of ASR. ...
Article
Full-text available
To obtain the correct analysis of electroencephalogram (EEG) signals, non-physiological and physiological artifacts should be removed from EEG signals. This study aims to give an overview on the existing methodology for removing physiological artifacts, e.g., ocular, cardiac, and muscle artifacts. The datasets, simulation platforms, and performance measures of artifact removal methods in previous related research are summarized. The advantages and disadvantages of each technique are discussed, including regression method, filtering method, blind source separation (BSS), wavelet transform (WT), empirical mode decomposition (EMD), singular spectrum analysis (SSA), and independent vector analysis (IVA). Also, the applications of hybrid approaches are presented, including discrete wavelet transform - adaptive filtering method (DWT-AFM), DWT-BSS, EMD-BSS, singular spectrum analysis - adaptive noise canceler (SSA-ANC), SSA-BSS, and EMD-IVA. Finally, a comparative analysis for these existing methods is provided based on their performance and merits. The result shows that hybrid methods can remove the artifacts more effectively than individual methods.
... An Artifacts Subspace Reconstruction (ASR) algorithm (Mullen et al., 2015) was applied to all retained channels to identify and remove transient or high-amplitude artifacts, e.g., eye movements. According to Chang and others (Chang et al., 2020), the application of this cleaning approach improves the accuracy of subsequent Independent Component Analysis (ICA). The ASR algorithm automatically identifies clean periods of the EEG data using a sliding window. ...
... According to the cut-off parameter k chosen by the user, the method can be more or less aggressive in removing EEG components portions of data. In this work, k = 20 was employed, according to the range suggested by Chang and others, who validated the ASR method on real EEG data and found that a k in the range 10-100 allows removing ocular artifacts still preserving cerebral components (Chang et al., 2020). Furthermore, ICA was applied to the ASR-cleaned EEG recordings exploiting the RUNICA Infomax algorithm as implemented in EEGLab (Makeig and Bell, 1996). ...
Article
Full-text available
Brain plasticity and functional reorganization are mechanisms behind functional motor recovery of patients after an ischemic stroke. The study of resting-state motor network functional connectivity by means of EEG proved to be useful in investigating changes occurring in the information flow and find correlation with motor function recovery. In the literature, most studies applying EEG to post-stroke patients investigated the undirected functional connectivity of interacting brain regions. Quite recently, works started to investigate the directionality of the connections and many approaches or features have been proposed, each of them being more suitable to describe different aspects, e.g., direct or indirect information flow between network nodes, the coupling strength or its characteristic oscillation frequency. Each work chose one specific measure, despite in literature there is not an agreed consensus, and the selection of the most appropriate measure is still an open issue. In an attempt to shed light on this methodological aspect, we propose here to combine the information of direct and indirect coupling provided by two frequency-domain measures based on Granger’s causality, i.e., the directed coherence (DC) and the generalized partial directed coherence (gPDC), to investigate the longitudinal changes of resting-state directed connectivity associated with sensorimotor rhythms α and β, occurring in 18 sub-acute ischemic stroke patients who followed a rehabilitation treatment. Our results showed a relevant role of the information flow through the pre-motor regions in the reorganization of the motor network after the rehabilitation in the sub-acute stage. In particular, DC highlighted an increase in intra-hemispheric coupling strength between pre-motor and primary motor areas, especially in ipsi-lesional hemisphere in both α and β frequency bands, whereas gPDC was more sensitive in the detection of those connection whose variation was mostly represented within the population. A decreased causal flow from contra-lesional premotor cortex towards supplementary motor area was detected in both α and β frequency bands and a significant reinforced inter-hemispheric connection from ipsi to contra-lesional pre-motor cortex was observed in β frequency. Interestingly, the connection from contra towards ipsilesional pre-motor area correlated with upper limb motor recovery in α band. The usage of two different measures of directed connectivity allowed a better comprehension of those coupling changes between brain motor regions, either direct or mediated, which mostly were influenced by the rehabilitation, revealing a particular involvement of the pre-motor areas in the cerebral functional reorganization.
... The weights depend on the similarity between blocks/patch. In the NLM algorithm, the similarity between two blocks (patches) is measured in terms of similarity between their neighborhoods as described in the equation (9). The weighting function (exponential as stated in Eq. 9) can be considered as a decaying function based on the similar patches found in the search neighborhood. ...
Preprint
Electroencephalogram (EEG) signals may get easily contaminated by muscle artifacts, which may lead to wrong interpretation in the brain-computer interface (BCI) system as well as in various medical diagnoses. The main objective of this paper is to remove muscle artifacts without distorting the information contained in the EEG. A novel multi-stage EEG denoising method is proposed for the first time where wavelet packet decomposition (WPD) is combined with modified non-local means (NLM) algorithm. At first, the artifacted EEG is identified through a pre-trained classifier. Next, the identified EEG signal is decomposed into wavelet coefficients through WPD. Muscle artifacts are eliminated from the wavelet coefficients by estimating the clean wavelet coefficients through a modified NLM algorithm instead of thresholding them. Finally, the artifact-free EEG is reconstructed from corrected wavelet coefficients through inverse WPD. To optimize the filter parameters of the NLM algorithm, two met-heuristic algorithms, grey wolf optimization (GWO) and particle swarm optimization (PSO), are used in this paper for the first time. The proposed system is first validated on simulated EEG data and then tested on real EEG data. The proposed approach achieved average mutual information (MI) as 2.9684 (0.7045) on real EEG data. The result reveals that the proposed system outperforms the recently developed denoising techniques with higher average MI, which indicates that the proposed approach is better in terms of quality of reconstruction and is fully automatic and can be implemented in online applications.
... In the second group, there are published papers, such as de Cheveigne and Arzounian [8] and Chang et al. [6]. In [8] authors have detected EEG and magnetoencephalography artefacts by their low correlation to other channels, and replaces them with the weighted sum of normal channels, a method called 'Inpainting' . ...
Article
Full-text available
Acquisition of neuronal signals involves a wide range of devices with specific electrical properties. Combined with other physiological sources within the body, the signals sensed by the devices are often distorted. Sometimes these distortions are visually identifiable, other times, they overlay with the signal characteristics making them very difficult to detect. To remove these distortions, the recordings are visually inspected and manually processed. However, this manual annotation process is time-consuming and automatic computational methods are needed to identify and remove these artefacts. Most of the existing artefact removal approaches rely on additional information from other recorded channels and fail when global artefacts are present or the affected channels constitute the majority of the recording system. Addressing this issue, this paper reports a novel channel-independent machine learning model to accurately identify and replace the artefactual segments present in the signals. Discarding these artifactual segments by the existing approaches causes discontinuities in the reproduced signals which may introduce errors in subsequent analyses. To avoid this, the proposed method predicts multiple values of the artefactual region using long–short term memory network to recreate the temporal and spectral properties of the recorded signal. The method has been tested on two open-access data sets and incorporated into the open-access SANTIA (SigMate Advanced: a Novel Tool for Identification of Artefacts in Neuronal Signals) toolbox for community use.
... cEEGrid data were analyzed using EEGLAB [30] (v.2019.1) on MATLAB (The Mathworks Inc.). Raw EEG data were down-sampled to 250 , high-pass filtered (1 ; Hamming window FIR, order: 826), cleaned for the 50 line noise (CleanLine function under EEGLAB; [66]) and denoised thanks to an Artifact Subspace Reconstruction (ASR; [67,49,68]) which removes artifacts based on a PCA. Finally, data were re-referenced to the average of R5 and L5 electrodes (closest to the mastoids, see Fig. 7.2b) and epoched time-locked to the sound stimuli (−1.5 to 2 around standard and odd sounds). ...
... Data were then re-referenced to a common average reference. Artifact subspace reconstruction (ASR) (Kothe and Jung, 2016;Mullen et al., 2015) with a cutoff parameter 20 was applied for automatic removal of large-amplitude artifacts like electrode pops and motion artifacts, which has been shown to improve subsequent ICA decomposition (Chang et al., 2019). ...
Article
This study applies adaptive mixture independent component analysis (AMICA) to learn a set of ICA models, each optimized by fitting a distributional model for each identified component process while maximizing component process independence within some subsets of time points of a multi-channel EEG dataset. Here, we applied 20-model AMICA decomposition to long-duration (1-2 hr), high-density (128-channel) EEG data recorded while participants used guided imagination to imagine situations stimulating the experience of 15 specified emotions. These decompositions tended to return models identifying spatiotemporal EEG patterns or states within single emotion imagination periods. Model probability transitions reflected time-courses of EEG dynamics during emotion imagination, which varied across emotions. Transitions between models accounting for imagined “grief” and “happiness” were more abrupt and better aligned with participant reports, while transitions for imagined “contentment” extended into adjoining “relaxation” periods. The spatial distributions of brain-localizable independent component processes (ICs) were more similar within participants (across emotions) than emotions (across participants). Across participants, brain regions with differences in IC spatial distributions (i.e., dipole density) between emotion imagination versus relaxation were identified in or near the left rostrolateral prefrontal, posterior cingulate cortex, right insula, bilateral sensorimotor, premotor, and associative visual cortex. No difference in dipole density was found between positive versus negative emotions. AMICA models of changes in high-density EEG dynamics may allow data-driven insights into brain dynamics during emotional experience, possibly enabling the improved performance of EEG-based emotion decoding and advancing our understanding of emotion.
... This function is based on artifact subspace reconstruction (ASR), which compares the structure of the artifactual EEG to that of known artifact-free reference data (Kothe and Jung, 2016). The tradeoff between artifactual and retaining brain activities depends on the ASR parameter, which we set to 20 according to the recommendations by (Chang et al., 2020). ...
Preprint
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 naturalistic viewing behavior. Here, we leverage simultaneous recording and analysis of eye movements and EEG to examine the formation of episodic memory in free viewing. We designed a stimulus screen with multiple elements that together comprised an event. Participants were asked to memorize several of these events while their EEG and eye movements were concurrently recorded. A subsequent cued-recall test assessed participants' memory for the element associations that specified each event. We overcame the problem of overlapping EEG responses to sequential saccades in free viewing using a deconvolution approach. We segmented EEG relative to the fixation onsets and examined EEG power in the theta and alpha frequency bands, the putative oscillatory correlates of episodic encoding mechanisms. We found that high subsequent memory performance was predicted by three modulations of fixation-related EEG: 1) theta synchronization at fixations after saccades relevant to event integration, 2) theta synchronization at fixations after within-element scrutinizing saccades, 3) alpha desynchronization at fixations after saccades between elements that were incidental to the task. 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 naturalistic viewing behavior.
... cEEGrid data were analyzed using EEGLAB [30] (v.2019.1) on MATLAB (The Mathworks Inc.). Raw EEG data were down-sampled to 250 , high-pass filtered (1 ; Hamming window FIR, order: 826), cleaned for the 50 line noise (CleanLine function under EEGLAB; [66]) and denoised thanks to an Artifact Subspace Reconstruction (ASR; [67,49,68]) which removes artifacts based on a PCA. Finally, data were re-referenced to the average of R5 and L5 electrodes (closest to the mastoids, see Fig. 7.2b) and epoched time-locked to the sound stimuli (−1.5 to 2 around standard and odd sounds). ...
Chapter
The objective of this chapter is to focus on the use of unobtrusive easy-to-use electrophysiological systems for neuroergonomic research. In a first section, we describe the challenges and limits related to the use of such systems. Electrode localization and signal processing solutions are then proposed to overcome some of the raised issues. In the second section, we explore the feasibility to measure pilots' auditory attention in a flight simulator using an unobtrusive EEG system [1] on a small number of participants. This study aims at measuring the cerebral activity associated with inattentional deafness in an ecological context with varying degrees of workload. The end goal was to assess the possibilities of such a system to be transferred to real-flight conditions. Our results illustrate this paradox: we were able to reproduce some of the results observed in the literature, but we also faced difficulties in terms of signal processing and measure identifications. We show that despite the lower signal-to-noise ratio observed with this kind of devices, we are able to detect event-related potentials (ERPs) and frequency features. In the last part, we discuss our results with regards to neuroadaptive systems challenges, and how we were able to overcome some of the current limits in neuroergonomics.
... According to the cut-off parameter k chosen by the user, the method can be more or less aggressive in removing EEG portions of data. In this work, a k = 15 was employed, accordingly to the range suggested by Chang et al., who validated the ASR method on real EEG data [20]. Furthermore, the independent component analysis (ICA) was applied to the ASR-cleaned EEG recordings exploiting the RUNICA Infomax algorithm as implemented in EEGLab [21]. ...
Article
Full-text available
Gait training in a virtual reality (VR) environment is promising for children affected by different disorders. However, the efficacy of VR therapy is still under debate, and more research is needed to clarify its effects on clinical conditions. The combination of VR with neuroimaging methods, such as the electroencephalography (EEG), might help in answering this need. The aim of the present work was to set up and test a system for the multimodal analysis of the gait pattern during VR gait training of pediatric populations by analyzing the EEG correlates as well as the kinematic and kinetic parameters of the gait. An EEG system was integrated with the Gait Real-time Analysis Interactive Lab (GRAIL). We developed and validated, with healthy adults (n = 5) and children (n = 4, healthy or affected by cerebral palsy (CP)), the hardware and software integration of the two systems, which allowed the synchronization of the acquired signals and a reliable identification of the initial contact (IC) of each gait cycle, showing good sensitivity and critical success index values. Moreover, we tested the multimodal acquisition by successfully analyzing EEG data and kinematic and kinetic parameters of one healthy child and one child with CP. This system gives the possibility of monitoring the effect of the VR therapy and studying the neural correlates of gait.
... The filtered EEG data was downsampled from 1000 to 250 Hz for efficient data analysis. As walking EEG signal is usually noisy and contaminated by large movement and muscle artifacts, two artifact removal methods, bad channel removal and artifact subspace reconstruction (ASR) [34,35], were applied to eliminate the noise. The channels were visually inspected to identify and remove broken or atypical channels. ...
Article
Full-text available
Brain stroke affects millions of people in the world every year, with 50 to 60 percent of stroke survivors suffering from functional disabilities, for which early and sustained post-stroke rehabilitation is highly recommended. However, approximately one third of stroke patients do not receive early in hospital rehabilitation programs due to insufficient medical facilities or lack of motivation. Gait triggered mixed reality (GTMR) is a cognitive-motor dual task with multisensory feedback tailored for lower-limb post-stroke rehabilitation, which we propose as a potential method for addressing these rehabilitation challenges. Simultaneous gait and EEG data from nine stroke patients was recorded and analyzed to assess the applicability of GTMR to different stroke patients, determine any impacts of GTMR on patients, and better understand brain dynamics as stroke patients perform different rehabilitation tasks. Walking cadence improved significantly for stroke patients and lower-limb movement induced alpha band power suppression during GTMR tasks. The brain dynamics and gait performance across different severities of stroke motor deficits was also assessed; the intensity of walking induced event related desynchronization (ERD) was found to be related to motor deficits, as classified by Brunnstrom stage. In particular, stronger lower-limb movement induced ERD during GTMR rehabilitation tasks was found for patients with moderate motor deficits (Brunnstrom stage IV). This investigation demonstrates the efficacy of the GTMR paradigm for enhancing lower-limb rehabilitation, explores the neural activities of cognitive-motor tasks in different stages of stroke, and highlights the potential for joining enhanced rehabilitation and real-time neural monitoring for superior stroke rehabilitation.
... The EEG signals were processed using a custom-built automated code including a Cleanline filter 48 , a finite impulse response filter between 3 and 40 Hz, the restoration of the reference-electrode FCz, re-referencing to common average, as well as a down-sampling to 256 Hz. In order to remove non-stereotype artifacts, the clean_rawdata EEGLAB plugin 49 was applied, which performs automated subspace reconstruction (ASR), a component-based method to effectively interpolate transient or large-amplitude artifacts 50 . After the application of ASR, an adaptive mixture independent component analysis (AMICA, Palmer 51 ) was applied to decompose the signal into brain and non-brain signals. ...
Article
Full-text available
The interaction of acute exercise and the central nervous system evokes increasing interest in interdisciplinary research fields of neuroscience. Novel approaches allow to monitor large‑scale brain networks from mobile electroencephalography (EEG) applying graph theory, but it is yet uncertain whether brain graphs extracted after exercise are reliable. We therefore aimed to investigate brain graph reliability extracted from resting state EEG data before and after submaximal exercise twice within one week in male participants. To obtain graph measures, we extracted global small‑world‑index (SWI), clustering coefficient (CC) and characteristic path length (PL) based on weighted phase leg index (wPLI) and spectral coherence (Coh) calculation. For reliability analysis, Intraclass‑Correlation‑Coefficient (ICC) and Coefficient of Variation (CoV) were computed for graph measures before (REST) and after POST) exercise. Overall results revealed poor to excellent measures at PRE and good to excellent ICCs at POST in the theta, alpha‑1 and alpha‑2, beta‑1 and beta‑2 frequency band. Based on bootstrap‑analysis, a positive effect of exercise on reliability of wPLI based measures was observed, while exercise induced a negative effect on reliability of Coh‑based graph measures. Findings indicate that brain graphs are a reliable tool to analyze brain networks in exercise contexts, which might be related to the neuroregulating effect of exercise inducing functional connections within the connectome. Relative and absolute reliability demonstrated good to excellent reliability after exercise. Chosen graph measures may not only allow analysis of acute, but also longitudinal studies in exercise‑scientific contexts.
... EEG data tend to be characterised by non-stationarity, a low signal-to-noise ratio, and artifacts from a number of sources including endogenous biological signals, background electrical interference, and variable electrode contact quality (33). In this study we implemented an automated data cleaning and pre-processing pipeline, using a variety of tools available through the EEGLAB toolbox in MATLAB (34), including: ICLabel using a trained algorithm to probabilistically label noise components from Independent Component Analysis (ICA) (35); Adaptive Mixture ICA (AMICA), which is demonstrated to be an effective variant of ICA for EEG artifact removal (36); and Artifact Subspace Reconstruction (ASR) (37). An automated approach was implemented to enable efficient and consistent pre-processing for EEG data across a larga number of participants, and to be consistent with the flexible and scaleable design approach of this analysis pipeline. ...
Preprint
Full-text available
Introduction To better understand the relationships between brain activity, cognitive function and mental health risk in adolescence there is value in identifying data-driven subgroups based on measurements of brain activity and function, and then comparing cognition and mental health symptoms between such subgroups. Methods Here we implement a multi-stage analysis pipeline to identify data-driven clusters of 12-year-olds (M = 12.64, SD = 0.32) based on frequency characteristics calculated from resting state, eyes-closed electroencephalography (EEG) recordings. EEG data was collected from 59 individuals as part of their baseline assessment in the Longitudinal Adolescent Brain Study (LABS) being undertaken in Queensland, Australia. Applying multiple unsupervised clustering algorithms to these EEG features, we identified well-separated subgroups of individuals. To study patterns of difference in cognitive function and mental health symptoms between core clusters, we applied Bayesian regression models to probabilistically identify differences in these measures between clusters. Results We identified 5 core clusters which were associated with distinct subtypes of resting state EEG frequency content. EEG features that were influential in differentiating clusters included Individual Alpha Frequency, relative power in 4 Hz bands up to 16 Hz, and 95% Spectral Edge Frequency. Bayesian models demonstrated substantial differences in psychological distress, sleep quality and cognitive function between these clusters. By examining associations between neurophysiology and health measures across clusters, we have identified preliminary risk and protective profiles linked to EEG characteristics. Conclusion In this work we have developed a flexible and scaleable pipeline to identify subgroups of individuals in early adolescence on the basis of resting state EEG activity. These findings provide new clues about neurophysiological subgroups of adolescents in the general population, and associated patterns of health and cognition that are not observed at the whole group level. This approach offers potential utility in clinical risk prediction for mental and cognitive health outcomes throughout adolescent development.
... We performed average referencing on EEG channels, then did artifact removal using artifact subspace reconstruction with the following parameters: Flatline Criterion = À1; Highpass = À1; ChannelCriterion = 0.6; LineNoiseCriterion = À1; BurstCriterion = 5; WindowCriterion = 0.25 (Christian Kothe, https://github.com/sccn/labstreaminglayer) (Chang et al., 2020). Independent component analysis (ICA; infomax binica) was then applied on the data to extract brain EEG sources as well as outside-brain artifact sources, such as blink, eye movement, and facial and neck muscle activation (Bell and Sejnowski, 1995;Makeig et al., 1996). ...
Article
Full-text available
Predicting and organizing patterns of events is important for humans to survive in a dynamically changing world. The motor system has been proposed to be actively, and necessarily, engaged in not only the production but the perception of rhythm by organizing hierarchical timing that influences auditory responses. It is not yet well understood how the motor system interacts with the auditory system to perceive and maintain hierarchical structure in time. This study investigated the dynamic interaction between auditory and motor functional sources during the perception and imagination of musical meters. We pursued this using a novel method combining high-density EEG, EMG and motion capture with independent component analysis (ICA) to separate motor and auditory activity during meter imagery while robustly controlling against covert movement. We demonstrated that endogenous brain activity in both auditory and motor functional sources reflects the imagination of binary and ternary meters in the absence of corresponding acoustic cues or overt movement at the meter rate. We found clear evidence for hypothesized motor-to-auditory information flow at the beat rate in all conditions, suggesting a role for top-down influence of the motor system on auditory processing of beat-based rhythms, and reflecting an auditory-motor system with tight reciprocal informational coupling. These findings align with and further extend a set of motor hypotheses from beat perception to hierarchical meter imagination, adding supporting evidence to active engagement of the motor system in auditory processing, which may more broadly speak to the neural mechanisms of temporal processing in other human cognitive functions.Significance StatementHumans live in a world full of hierarchically structured temporal information, the accurate perception of which is essential for understanding speech and music. Music provides a window into the brain mechanisms of time perception, enabling us to examine how the brain groups musical beats into, for example a march or waltz. Using a novel paradigm combining measurement of electrical brain activity with data-driven analysis, this study directly investigates motor-auditory connectivity during meter imagination. Findings highlight the importance of the motor system in the active imagination of meter. This study sheds new light on a fundamental form of perception by demonstrating how auditory-motor interaction may support hierarchical timing processing, which may have clinical implications for speech and motor rehabilitation.
... and visual inspection (C. Y. Chang et al., 2019;Plechawska-Wojcik et al., 2018). Independent component analysis (ICA) using the ICA toolbox with extended infomax set to 1 was applied to the resulting signal data for all electrode sources to classify signal variance associated with vertical and horizontal eye-blinks and heart rate (where applicable), with a maximum of 12 components. ...
Article
Full-text available
Collaboration between two individuals is thought to be associated with the synchrony of two different brain activities. Indeed, prefrontal cortical activation and alpha frequency band modulation has been widely reported, but little is known about interbrain synchrony (IBS) changes occurring during social interaction such as collaboration or competition. In this study, we assess the dynamic of IBS variation in order to provide novel insights into the frequency band modulation underlying collaboration. To address this question, we used electroencephalography (EEG) to simultaneously record the brain activity of two individuals playing a computer‐based game facing four different conditions: collaboration, competition, single participation, and passive observation. The computer‐based game consisted of a fast button response task. Using data recorded in sensor space, we calculated an IBS value for each frequency band using both wavelet coherence transform and phase‐locking value and performed single‐subject analysis to compare each condition. We found significant IBS in frontal electrodes only present during collaboration associated with alpha frequency band modulation. In addition, we observed significant IBS in the theta frequency band for both collaboration and competition conditions, along with a significant single‐subject cortical activity. Competition is distinguishable through single‐subject activity in several regions and frequency bands of the brain. Performance is correlated with single‐subject frontal activation during collaboration in the alpha and beta frequency band.
... Then, data were 1 Hz high-pass filtered 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). ...
Preprint
Full-text available
The pronunciation of foreign phonemes is assumed to involve auditory feedback control processes that compare vocalized phonemes to target sounds. The electrophysiological correlate of this process is known as the speaking-induced suppression (SIS) of early auditory evoked activity. To gain insight into the neural processes that mediate the learning of foreign phoneme pronunciation, we recorded event-related potentials (ERP) when participants (N=19) pronounced either native or foreign phonemes. Analyses of single-trial ERPs revealed no differences in SIS between foreign and native phonemes in early time-windows (approx. 85-290 ms). In contrast, the amplitude of the fronto-centrally distributed late slow wave (LSW, 320-440 ms) was modulated by the pronunciation of foreign phonemes. Whereas the self-produced native phonemes evoked a constant amplitude LSW, the LSW evoked by self-vocalized foreign phonemes shifted towards more positive amplitudes across the experiment. Importantly, the LSW amplitude correlated positively with the improved pronunciation of the foreign phoneme. These results suggest that the LSW may reflect higher-order internal monitoring processes that signal successful pronunciation and enable adjustments to future vocalization.
... The value used in the arg_burst parameter of clean raw_data function (5 standard deviations) during the EEG preprocessing was based on the parameters suggested in Mullen et al., 2015. However, new empirical results (Chang et al., 2020) suggest that a number in the range of 20-30 standard deviations should be used. 5 SD can be aggressive and lead to loss of brain information. ...
Book
Full-text available
The main objective of this Research Topic was to gather studies that shed more light on the benefits of physical exercise in the neurophysiological system, from childhood to old age and from the field of health to sports or professional performance. For example, we consider important those studies that deepen into the epigenetic mechanisms involved in the aging process and their modulation through physical exercise, improving prevention and treatment therapies, and those that contributes to better understand how physical activity improves brain functions (e.g., increased hippocampal), or what effect cognitive loads cause in variables such as heart rate variability or brain waves. We also consider it particularly interesting to show studies that can reflect how physical exercise can be a good preventive strategy to avoid or counteract neurodegenerative diseases, such as Alzheimer, and consequently, increase the time and quality of life. Thus, some of the topics of interest for this Research Topic are studies that contemplate the latest advances on neurophysiological and epigenetic effects of physical exercise on the aging, or beneficial effects of the practice of physical activity and sport on anti-aging and neuroprotective mechanisms. Equally relevant aspects to consider are the effects of physical exercise to prevent neurodegenerative diseases, the relationship between physical exercise practice and improvement of brain functions, the effects of cognitive loads at the neurophysiological level, or the neurophysiological system behavior related to sports or professional performance.
... An artifact subspace reconstruction approach was carried out with the clean_rawdata function (with default parameters) to repair data segments of an artifact by applying a reconstruction mixing matrix from non-interpolated neighboring channels. The mixing matrix is computed from clean segments within the EEG data 90 . Blind source separation was performed with temporal Independent Component Analysis on each preprocessed dataset using the extended INFOMAX algorithm 91,92 with principal component analysis rank reduction (further reduced for interpolated channels). ...
Article
Full-text available
Fragile X Syndrome (FXS) is a monogenetic form of intellectual disability and autism in which well-established knockout (KO) animal models point to neuronal hyperexcitability and abnormal gamma-frequency physiology as a basis for key disorder features. Translating these findings into patients may identify tractable treatment targets. Using source modeling of resting-state electroencephalography data, we report findings in FXS, including 1) increases in localized gamma activity, 2) pervasive changes of theta/alpha activity, indicative of disrupted thalamocortical modulation coupled with elevated gamma power, 3) stepwise moderation of low and high-frequency abnormalities based on female sex, and 4) relationship of this physiology to intellectual disability and neuropsychiatric symptoms. Our observations extend findings in Fmr1 −/− KO mice to patients with FXS and raise a key role for disrupted thalamocortical modulation in local hyperexcitability. This systems-level mechanism has received limited preclinical attention but has implications for understanding fundamental disease mechanisms.
... It can be further expanded in terms of dataset. Third, some classical denoising methods were employed in this work, but many denoising methods with better effect were now applied, like Artifact subspace reconstruction (ASR) [47], morphological component analysis (MCA) [48], the surrogatebased artifact removal (SuBAR) [49]. These methods can be tried in the future work for the denoising EEG signal recorded during sleep. ...
Article
Elimination of intra-artifacts in EEG has been overlooked in most of the existing sleep staging systems, especially in deep learning-based approaches. Whether intra-artifacts, originated from the eye movement, chin muscle firing, or heart beating, etc., in EEG signals would lead to a positive or a negative masking effect on deep learning-based sleep staging systems was investigated in this paper. We systematically analyzed several traditional pre-processing methods involving fast Independent Component Analysis (FastICA), Information Maximization (Infomax), and Second-order Blind Source Separation (SOBI). On top of these methods, a SOBI-WT method based on the joint use of the SOBI and Wavelet Transform (WT) is proposed. It offered an effective solution for suppressing artifact components while retaining residual informative data. To provide a comprehensive comparative analysis, these pre-processing methods were applied to eliminate the intra-artifacts and the processed signals were fed to two ready-to-use deep learning models, namely two-step hierarchical neural network (THNN) and SimpleSleepNet for automatic sleep staging. The evaluation was performed on two widely used public datasets, Montreal Archive of Sleep Studies (MASS) and Sleep-EDF Expanded, and a clinical dataset that was collected in Huashan Hospital of Fudan University, Shanghai, China (HSFU). The proposed SOBI-WT method increased the accuracy from 79.0% to 81.3% on MASS, 83.3% to 85.7% on Sleep-EDF Expanded, and 75.5% to 77.1% on HSFU compared with the raw EEG signal, respectively. Experimental results demonstrate that the intra-artifacts bring out a masking negative impact on the deep learning-based sleep staging systems and the proposed SOBI-WT method has the best performance in diminishing this negative impact compared with other artifact elimination methods.
... improved automated Wavelet-ICA (EAWICA) [26], hybrid ICA-Wavelet rework technique (ICA-W) [27] or through growing new approaches https://www.linkedin.com/pulse/brain-computer-interfaces-bci-robotics-user-qamar-ul-islam/ 9/26 along with adaptive de-noising frameworks [28] and Artifact Subspace Reconstruction (ASR) [29]. feature extraction includes recognising useful information (e.g., spectral strength, time epochs, spatial filtering) for higher discriminability among intellectual states. ...
Method
Full-text available
Brain computer interface (BCI) systems set up a direct communication among the brain and an outside device. those systems may be used for amusement, to improve the quality of lifestyles of patients and to manipulate virtual and augmented reality applications, industrial machines, and robots. inside the neuroscience discipline along with in neurorehabilitation, BCIs are incorporated into controlled virtual environments used for the treatment of disability or cognitive development of subjects, as an instance, in case of cerebral palsy, Down syndrome, and melancholy. Its goal is to promote a recovery of mind function misplaced because of a lesion through noninvasive brain stimulation (brain modulation) in a more accurate and quicker way than the conventional strategies. Neurorobotics combines BCIs with robotics aiming to broaden artificial limbs, that may act as actual members of human body being managed from a brain-machine interface. With the development of a better understanding of how our mind works, new sensible computational algorithms are being taken into consideration, making it feasible to simulate and version-specific brain functions for the improvement of recent Computational Intelligence algorithms and, eventually, BCI for cellular gadgets/apps. For more info: https://www.linkedin.com/pulse/brain-computer-interfaces-bci-robotics-user-qamar-ul-islam/
... The raw EEG signals were first down-sampled to 250 Hz and filtered with a band pass filter of 1.5-70 Hz. After this initial filtering step, line noise and other large non-stationary artifacts were identified and cleaned using the artifact subspace reconstruction (ASR) approach (Chang et al., 2020). The cleaned signals were re-referenced to a common average reference. ...
Article
Full-text available
The flow state-an experience of complete absorption in an activity-is linked with less self-referential processing and increased arousal. We used the heart-evoked potential (HEP), an index representing brain-heart interaction, as well as indices of peripheral physiology to assess the state of flow in individuals playing a video game. 22 gamers and 21 non-gamers played the video game Thumper for 25 min while their brain and cardiorespiratory signals were simultaneously recorded. The more participants were absorbed in the game, the less they thought about time and the faster time passed subjectively. On the cortical level, the fronto-central HEP amplitude was significantly lower while playing the game compared to resting states before and after the game, reflecting less self-referential processing while playing. This HEP effect corresponded with lower activity during gameplay in brain regions contributing to interoceptive processing. The HEP amplitude predicted the level of absorption in the game. While the HEP amplitude was overall lower during the gaming session than during the resting states, within the gaming session the amplitude of HEP was positively associated with absorption. Since higher absorption was related to higher performance in the game, the higher HEP in more absorbed individuals reflects more efficient brain-heart interaction, which is necessary for efficient game play. On the physiological level, a higher level of flow was associated with increased overall sympathetic activity and less inhibited parasympathetic activity toward the end of the game. These results are building blocks for future neurophysiological assessments of flow.
... Artifact subspace reconstruction (ASR) [81,82] was carried out through its EEGLAB implementation in the function clean_artifacts. This function was used to remove channels that showed no signal activity (flat line threshold: 10 s), noisy signals (noisy line threshold: 4 std), or a poor correlation with adjacent channels (correlation threshold: 0.75). ...
Article
This study investigated the neural dynamics associated with short-term exposure to different virtual classroom designs with different window placement and room dimension. Participants engaged in five brief cognitive tasks in each design condition including the Stroop Test, the Digit Span Test, the Benton Test, a Visual Memory Test, and an Arithmetic Test. Performance on the cognitive tests and Electroencephalogram (EEG) data were analyzed by contrasting various classroom design conditions. The cognitive-test-performance results showed no significant differences related to the architectural design features studied. We computed frequency band-power and connectivity EEG features to identify neural patterns associated to environmental conditions. A leave-one-out machine-learning classification scheme was implemented to assess the robustness of the EEG features, with the classification accuracy evaluation of the trained model repeatedly performed against an unseen participant’s data. The classification results located consistent differences in the EEG features across participants in the different classroom design conditions, with a predictive power (test-set accuracy: 51.5%-61.3%) that was significantly higher compared to a baseline classification learning outcome using scrambled data. These findings were most robust during the Visual Memory Test, and were not found during the Stroop Test and the Arithmetic Test. The most discriminative EEG features were observed in bilateral occipital, parietal, and frontal regions in the theta (4-8 Hz) and alpha (8-12 Hz) frequency bands. Connectivity analysis reinforced these findings by showing that there were changes in the transfer of information from centro-parietal to frontal electrodes in the different classroom conditions. While the implications of these findings for student learning are yet to be determined, this study provides rigorous evidence that brain activity features during cognitive tasks are affected by the design elements of window placement and room dimensions. The ongoing development of this EEG-based approach has the potential to strengthen evidence-based design through the use of solid neurophysiological evidence.
... It is therefore of paramount importance that EEG artifacts are properly detected so as not to induce false alarms in epilepsy detection systems. After detecting an artifact successfully, the epilepsy detection system could then intelligently decide whether to adaptively correct the EEG signal (using methods such as Artifact Subspace Reconstruction [13]), or to extract features that can help the system classify upcoming seizures or disregard the data. To this end, there is a need for estimations of the generic presence of artifacts (binary classification, BC), if not a per-channel estimation of the presence of artifacts (multi-label classification, MC), or even better, an estimation of the specific source cause of artifacts on each channel (multiclass multi-output classification, MMC). ...
Preprint
Full-text available
In the context of epilepsy monitoring, EEG artifacts are often mistaken for seizures due to their morphological similarity in both amplitude and frequency, making seizure detection systems susceptible to higher false alarm rates. In this work we present the implementation of an artifact detection algorithm based on a minimal number of EEG channels on a parallel ultra-low-power (PULP) embedded platform. The analyses are based on the TUH EEG Artifact Corpus dataset and focus on the temporal electrodes. First, we extract optimal feature models in the frequency domain using an automated machine learning framework, achieving a 93.95% accuracy, with a 0.838 F1 score for a 4 temporal EEG channel setup. The achieved accuracy levels surpass state-of-the-art by nearly 20%. Then, these algorithms are parallelized and optimized for a PULP platform, achieving a 5.21 times improvement of energy-efficient compared to state-of-the-art low-power implementations of artifact detection frameworks. Combining this model with a low-power seizure detection algorithm would allow for 300h of continuous monitoring on a 300 mAh battery in a wearable form factor and power budget. These results pave the way for implementing affordable, wearable, long-term epilepsy monitoring solutions with low false-positive rates and high sensitivity, meeting both patients' and caregivers' requirements.
... In early research, approaches based on blind source separation (BSS) were explored for the automatic removal of EMG artifacts from multichannel EEG [8]. The authors in [9][10][11] showed that independent component analysis (ICA) resulted in a good performance for removing EMG artifacts from the multichannel EEG data. Janani et al. [12] showed that canonical correlation analysis (CCA) outperformed the ICA to eliminate EMG from EEG successfully. ...
Article
Full-text available
Electroencephalogram (EEG) signals may get easily contaminated by muscle artifacts, which may lead to wrong interpretation in the brain–computer interface (BCI) system as well as in various medical diagnoses. The main objective of this paper is to remove muscle artifacts without distorting the information contained in the EEG. A novel multi-stage EEG denoising method is proposed for the first time in which wavelet packet decomposition (WPD) is combined with a modified non-local means (NLM) algorithm. At first, the artifact EEG signal is identified through a pre-trained classifier. Next, the identified EEG signal is decomposed into wavelet coefficients and corrected through a modified NLM filter. Finally, the artifact-free EEG is reconstructed from corrected wavelet coefficients through inverse WPD. To optimize the filter parameters, two meta-heuristic algorithms are used in this paper for the first time. The proposed system is first validated on simulated EEG data and then tested on real EEG data. The proposed approach achieved average mutual information (MI) as 2.9684 ± 0.7045 on real EEG data. The result reveals that the proposed system outperforms recently developed denoising techniques with higher average MI, which indicates that the proposed approach is better in terms of quality of reconstruction and is fully automatic.
... Continuous EEG data were filtered via FIR filter between 0.5 and 40 Hz using Hamming window with the help of the pop_eegfiltnew function in EEGLAB. After that, an automatic method known as artifact subspace reconstruction (ASR) is used to reconstruct the artifacts portion of the data with clean data [44]. This method removes the non-stationary high variance signal and then reconstructs the missing data using a spatial mixing matrix. ...
Article
Full-text available
This study investigates the effect of long-term alpha neurofeedback training (NFT) in healthy adults using music stimuli. The optimal protocol for future research is presented in this study. The data from 40 healthy participants, divided into two groups (NFT group and Control group), were analyzed in the current study. We found a significantly enhanced alpha rhythm after training in the NFT group which was not observed in the control group. The immediate subsequent effects were greater in more than 80% of the sessions from the initial recordings. Stroop task and behavioral questionnaires, mini-mental state exam (MMSE), and perceived stress scale (PSS) did not reveal any training-specific changes. Within-training session effects were significant from the baseline and were more pronounced at the beginning of the session as compared to the end of the session. It is also observed that a shorter session length with multiple sessions may be more effective than a long and continuous run of a single session.
... 9) ICA_subtract, which is probably the most commonly used approach in EEG researchthis pipeline computed ICA, then subtracted the components identified as artifacts from the ICA unmixing matrix using ICLabel, then reconstructed the electrode space data (without any MWF cleaning applied) (Pion-Tonachini et al., 2019). 10) ASR applied the Artifact Subspace Reconstruction (ASR) approach followed by ICA subtraction of artifacts identified by ICLabel (Chang et al., 2019). 11) HAPPE applied the Harvard Automated Processing Pipeline for EEG (Gabard-Durnam et al., 2018). ...
Preprint
Full-text available
Electroencephalographic (EEG) data is typically contaminated with non-neural artifacts which can confound the results of experiments. Artifact cleaning approaches are available, but often require time-consuming manual input and significant expertise. Advancements in artifact cleaning often only address a single artifact, are only compared against a small selection of pre-existing methods, and seldom assess whether a proposed advancement improves experimental outcomes. To address these issues, we developed RELAX (the Reduction of Electroencephalographic Artifacts), an automated EEG cleaning pipeline implemented within EEGLAB that reduces all artifact types. RELAX cleans continuous data using Multiple Wiener filtering [MWF] and/or wavelet enhanced independent component analysis [wICA] applied to artifacts identified by ICLabel [wICA_ICLabel]). Several versions of RELAX were tested using three datasets containing a mix of cognitive and resting recordings (N = 213, 60 and 23 respectively). RELAX was compared against six commonly used EEG cleaning approaches across a wide range of artifact cleaning quality metrics, including signal-to-error and artifact-to-residue ratios, measures of remaining blink and muscle activity, and the amount of variance explained by experimental manipulations after cleaning. RELAX with MWF and wICA_ICLabel showed amongst the best performance for cleaning blink and muscle artifacts while still preserving neural signal. RELAX with wICA_ICLabel (and no MWF) may perform better at detecting the effect of experimental manipulations on alpha oscillations in working memory tasks. The pipeline is easy to implement in MATLAB and freely available on GitHub. Given its high cleaning performance, objectivity, and ease of use, we recommend RELAX for data cleaning across EEG studies.
... The noisy bursts in the filtered EEG data were corrected with Artifact Subspace Reconstruction (ASR) approach, which is an EEGLAB plug-in (Mullen et al., 2015). The threshold for burst detection criterion (specified as k in the ASR algorithm) was set as 20 as recommended in Chang et al. (2019) based on a rigorous evaluation. Afterwards, the EEG signals were re-referenced to the common average, i.e., the arithmetic average of all electrodes (Nunez and Srinivasan, 2006), which is advocated for EEG source estimation. ...
Article
Full-text available
Repeatedly performing a submaximal motor task for a prolonged period of time leads to muscle fatigue comprising a central and peripheral component, which demands a gradually increasing effort. However, the brain contribution to the enhancement of effort to cope with progressing fatigue lacks a complete understanding. The intermittent motor tasks (IMTs) closely resemble many activities of daily living (ADL), thus remaining physiologically relevant to study fatigue. The scope of this study is therefore to investigate the EEG-based brain activation patterns in healthy subjects performing IMT until self-perceived exhaustion. Fourteen participants (median age 51.5 years; age range 26−72 years; 6 males) repeated elbow flexion contractions at 40% maximum voluntary contraction by following visual cues displayed on an oscilloscope screen until subjective exhaustion. Each contraction lasted ≈5 s with a 2-s rest between trials. The force, EEG, and surface EMG (from elbow joint muscles) data were simultaneously collected. After preprocessing, we selected a subset of trials at the beginning, middle, and end of the study session representing brain activities germane to mild, moderate, and severe fatigue conditions, respectively, to compare and contrast the changes in the EEG time-frequency (TF) characteristics across the conditions. The outcome of channel- and source-level TF analyses reveals that the theta, alpha, and beta power spectral densities vary in proportion to fatigue levels in cortical motor areas. We observed a statistically significant change in the band-specific spectral power in relation to the graded fatigue from both the steady- and post-contraction EEG data. The findings would enhance our understanding on the etiology and physiology of voluntary motor-action-related fatigue and provide pointers to counteract the perception of muscle weakness and lack of motor endurance associated with ADL. The study outcome would help rationalize why certain patients experience exacerbated fatigue while carrying out mundane tasks, evaluate how clinical conditions such as neurological disorders and cancer treatment alter neural mechanisms underlying fatigue in future studies, and develop therapeutic strategies for restoring the patients' ability to participate in ADL by mitigating the central and muscle fatigue.
... Jan et al. [20] improved the ICA method to better artifact removal. Chang et al. [21] used artifact subspace reconstruction (ASR) to preprocess EEG data and, combined with the ICA separation method, greatly improved the accuracy of artifact removal. Indirect separation of artifacts avoids extra electrodes, making it convenient in the test environment and reducing the extra noise. ...
Article
Full-text available
As a convenient device for observing neural activity in the natural environment, portable EEG technology (PEEGT) has an extensive prospect in expanding neuroscience research into natural applications. However, unlike in the laboratory environment, PEEGT is usually applied in a semiconstrained environment, including management and engineering, generating much more artifacts caused by the subjects’ activities. Due to the limitations of existing artifacts annotation, the problem limits PEEGT to take advantage of portability and low-test cost, which is a crucial obstacle for the potential application of PEEGT in the natural environment. This paper proposes an intelligent method to identify two leading antecedent causes of EEG artifacts, participant’s blinks and head movements, and annotate the time segments of artifacts in real time based on computer vision (CV). Furthermore, it changes the original postprocessing mode based on artifact signal recognition to the preprocessing mode based on artifact behavior recognition by the CV method. Through a comparative experiment with three artifacts mark operators and the CV method, we verify the effectiveness of the method, which lays a foundation for accurate artifact removal in real time in the next step. It enlightens us on how to adopt computer technology to conduct large-scale neurotesting in a natural semiconstrained environment outside the laboratory without expensive laboratory equipment or high manual costs.
Conference Paper
Brain-computer interface (BCI) systems allow users to communicate directly with a device using their brain. BCI devices leveraging electroencephalography (EEG) signals as a means of communication typically use manual feature engineering on the data to perform decoding. This approach is time intensive, requires substantial domain knowledge, and does not translate well, even to similar tasks. To combat this issue, we designed a convolutional neural network (CNN) model to perform decoding on EEG data collected from an auditory attention paradigm. Our CNN model not only bypasses the need for manual feature engineering, but additionally improves decoding accuracy (∼77%) and efficiency (∼11 bits/min) compared to a support vector machine (SVM) baseline. The results demonstrate the potential for the use of CNN in auditory BCI designs.
Conference Paper
Combining electroencephalography (EEG) to functional near-infrared spectroscopy (fNIRS) is a promising technique that has gained momentum thanks to their complementarity. While EEG measures the electrical activity of the brain, fNIRS records the variations in cerebral blood flow and related hemoglobin concentrations. However, both modalities are typically contaminated with artefacts. Muscle and eye artefacts, affect the EEG signals, while hemodynamic and oxygenation changes in the extracerebral compartment due to systemic changes (superficial layer) corrupt the fNIRS signals. Moreover, both signals are sensitive to sensor motion artefacts characterized by large amplitude. There are several well-established methods for removing artefacts for both modalities. The objective of this paper is to apply a common approach to denoise both EEG and fNIRS signals. Indeed Artifact Subspace Reconstruction (ASR) method, which is an automatic, online-capable and efficient method for deleting transient or large-amplitude EEG artefacts, can be a good alternative to also denoise fNIRS signals. In this paper, we first propose, a new more comprehensive formulation of ASR. Then, we study the effectiveness of the method in denoising both the EEG and fNIRS signals.
Article
Lay abstract: This study investigates the effects of a probiotic on preschoolers' brain electrical activity with autism spectrum disorder. Autism is a disorder with an increasing prevalence characterized by an enormous individual, family, and social cost. Although the etiology of autism spectrum disorder is unknown, an interaction between genetic and environmental factors is implicated, converging in altered brain synaptogenesis and, therefore, connectivity. Besides deepening the knowledge on the resting brain electrical activity that characterizes this disorder, this study allows analyzing the positive central effects of a 6-month therapy with a probiotic through a randomized, double-blind placebo-controlled study and the correlations between electroencephalography activity and biochemical and clinical parameters. In subjects treated with probiotics, we observed a decrease of power in frontopolar regions in beta and gamma bands, and increased coherence in the same bands together with a shift in frontal asymmetry, which suggests a modification toward a typical brain activity. Electroencephalography measures were significantly correlated with clinical and biochemical measures. These findings support the importance of further investigations on probiotics' benefits in autism spectrum disorder to better elucidate mechanistic links between probiotics supplementation and changes in brain activity.
Article
Full-text available
Mental state changes induced by stimuli under experimental settings or by daily events in real life affect task performance and are entwined with physical and mental health. In this study, we developed a physiological state indicator with five parameters that reflect the subject’s real-time physiological states based on online EEG signal processing. These five parameters are attention, fatigue, stress, and the brain activity shifts of the left and right hemispheres. We designed a target detection experiment modified by a cognitive attention network test for validating the effectiveness of the proposed indicator, as such conditions would better approximate a real chaotic environment. Results demonstrated that attention levels while performing the target detection task were significantly higher than during rest periods, but also exhibited a decay over time. In contrast, the fatigue level increased gradually and plateaued by the third rest period. Similar to attention levels, the stress level decreased as the experiment proceeded. These parameters are therefore shown to be highly correlated to different stages of the experiment, suggesting their usage as primary factors in passive brain-computer interfaces (BCI). In addition, the left and right brain activity indexes reveal the EEG neural modulations of the corresponding hemispheres, which set a feasible reference of activation for an active BCI control system, such as one executing motor imagery tasks. The proposed indicator is applicable to potential passive and active BCI applications for monitoring the subject’s physiological state change in real-time, along with providing a means of evaluating the associated signal quality to enhance the BCI performance.
Article
An auditory-visual speech benefit, the benefit that visual speech cues bring to auditory speech perception, is experienced from early on in infancy and continues to be experienced to an increasing degree with age. While there is both behavioural and neurophysiological evidence for children and adults, only behavioural evidence exists for infants – as no neurophysiological study has provided a comprehensive examination of the auditory-visual speech benefit in infants. It is also surprising that most studies on auditory-visual speech benefit do not concurrently report looking behaviour especially since the auditory-visual speech benefit rests on the assumption that listeners attend to a speaker's talking face and that there are meaningful individual differences in looking behaviour. To address these gaps, we simultaneously recorded electroencephalographic (EEG) and eye-tracking data of 5-month-olds, 4-year-olds and adults as they were presented with a speaker in auditory-only (AO), visual-only (VO), and auditory-visual (AV) modes. Cortical tracking analyses that involved forward encoding models of the speech envelope revealed that there was an auditory-visual speech benefit [i.e., AV > (A+V)], evident in 5-month-olds and adults but not 4-year-olds. Examination of cortical tracking accuracy in relation to looking behaviour, showed that infants’ relative attention to the speaker's mouth (vs. eyes) was positively correlated with cortical tracking accuracy of VO speech, whereas adults’ attention to the display overall was negatively correlated with cortical tracking accuracy of VO speech. This study provides the first neurophysiological evidence of auditory-visual speech benefit in infants and our results suggest ways in which current models of speech processing can be fine-tuned.
Article
Background The evidences for demonstrating the contributions of the cerebral cortex in human postural control is increasing. However, there remain little insights about the cortical correlates of balance control in lower-limb amputees. The present study aimed to investigate the cortical activity and balance performance of transfemoral amputees in comparison to healthy individuals during a continuous balance task (CBT). Methods The postural stability of the participants was defined with limit of stability parameter. Electroencephalography (EEG) data were recorded in synchronization with the center of pressure (CoP) data from eighteen individuals (including eight unilateral transfemoral amputees). We anticipated that, due to the limb loss, the postural demand of transfemoral amputees increases which significantly modulates the spectral power of intrinsic cortical oscillations. Findings Using the independent components from the sensorimotor areas and supplementary motor area (SMA), our results present a well-pronounced drop of alpha spectral power at sensorimotor area contralateral to sound limb of amputees in comparison to SMA and the sensorimotor area contralateral to prosthetic limb. Following this, we found significantly higher (p < 0.05) limit of stability (LOS) at their sound limb than at the prosthetic limb. Healthy individuals have similar contribution from both the limbs and the EEG alpha spectral power was similar across the three regions of the cortex during the balance control task as expected. Overall, a decent correlation was found between the LOS and alpha spectral power in both amputee and healthy individuals (Pearson’s correlation coefficient > 0.5). Interpretation By externally stimulating the highlighted cortical regions, neuroplasticity might be promoted which helps to reduce the training time for the efficient rehabilitation of amputees. Additionally, this new knowledge might benefit in the designing and development of innovative interventions to prevent falls due to lower limb amputation.
Article
Objective Delta-gamma phase-amplitude coupling in EEG is useful for localizing epileptic sources and evaluate severity in children with infantile spasms. We (1) develop an automated EEG preprocessing pipeline to clean data using artifact subspace reconstruction (ASR) and independent component (IC) analysis (ICA) and (2) evaluate delta-gamma modulation index (MI) as a method to distinguish children with epileptic spasms (cases) from normal controls during sleep and awake. Methods Using 400 scalp EEG datasets (200 sleep, 200 awake) from 100 subjects, we calculated MI after applying high-pass and line-noise filters (Clean 0), and after ASR followed by either conservative (Clean 1) or stringent (Clean 2) artifactual IC rejection. Classification of cases and controls using MI was evaluated with Receiver Operating Characteristics (ROC) to obtain area under curve (AUC). Results The artifact rejection algorithm reduced raw signal variance by 29-45% and 38-60% for Clean 1 and Clean 2, respectively. MI derived from sleep data, with or without preprocessing, robustly classified the groups (all AUC > 0.98). In contrast, group classification using MI derived from awake data was successful only after Clean 2 (AUC = 0.85). Conclusions We have developed an automated EEG preprocessing pipeline to perform artifact rejection and quantify delta-gamma modulation index.
Article
Full-text available
Brain connectivity can be estimated through many analyses applied to electroencephalographic (EEG) data. However, substantial heterogeneity in the implementation of connectivity methods exist. Heterogeneity in conceptualization of connectivity measures, data collection, or data pre-processing may be associated with variability in robustness of measurement. While it is difficult to compare the results of studies using different EEG connectivity measures, standardization of processing and reporting may facilitate the task. We discuss how factors such as referencing, epoch length and number, controls for volume conduction, artefact removal, and statistical control of multiple comparisons influence the EEG connectivity estimate for connectivity measures, and what can be done to control for potential confounds associated with these factors. Based on the results reported in previous literature, this article presents recommendations and a novel checklist developed for quality assessment of EEG connectivity studies. This checklist and its recommendations are made in an effort to draw attention to factors that may influence connectivity estimates and factors that need to be improved in future research. Standardization of procedures and reporting in EEG connectivity may lead to EEG connectivity studies to be made more synthesisable and comparable despite variations in the methodology underlying connectivity estimates.
Article
We aimed to test the idea that rhythmic transcranial magnetic stimulation (TMS) entrains cortical oscillations. To do this, we examined oscillatory responses in the electroencephalogram (EEG) to TMS over primary motor cortex. In particular, we contrasted responses to real TMS with those to sham TMS in order to dissociate the contributions of direct (transcranial) activation and indirect activation (via auditory/sensory input) of the brain. We first showed that real single pulse TMS elicited a brief (∼200 ms) increase in sensorimotor beta power whose frequency closely matched that of each individual's post-movement beta rebound (PMBR, ∼18 Hz). Sham TMS triggered minimal oscillatory activity. Together this implies that real TMS interacts with endogenous oscillations via direct brain activation. We then showed that although trains of real rhythmic TMS delivered at each individuals PMBR frequency produced a brief increase in beta power at the same frequency, real arrhythmic TMS also elicited an equivalent increase in beta. The implication is that the oscillatory response is independent of the rhythm of stimulation. By contrast, sham stimulation elicited minimal activity in the beta band, and the responses to rhythmic and arrhythmic sham TMS were broadly similar, showing that sham rhythmic stimulation did not produce entrainment via sensory rhythms. Together, the data demonstrate that the beta oscillatory response of M1 to real TMS predominantly reflects direct activation of the underlying cortex. However, the data do not support the notion of rhythmic TMS enhancing oscillatory activity via entrainment-like mechanisms, at least within the constraints of the current experimental set-up.
Article
Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder that pervasively interferes with the lives of individuals starting in childhood. Objective: To address the subjectivity of current diagnostic approaches, many studies have been dedicated to efforts to identify the differences between ADHD and neurotypical (NT) individuals using EEG and continuous performance tests (CPT). Approach: In this study, we proposed EEG-based long short-term memory (LSTM) networks that utilize deep learning techniques with learning the cognitive state transition to discriminate between ADHD and NT children via EEG signal processing. A total of thirty neurotypical children and thirty ADHD children participated in CPT tests while being monitored with EEG. Several architectures of deep and machine learning were applied to three EEG data segments including resting state, cognitive execution, and a period containing a fusion of those. Main results: The experimental results indicated that EEG-based LSTM networks produced the best performance with an average accuracy of 90.50 ± 0.81 % in comparison with the deep neural networks, the convolutional neural networks, and the support vector machines with learning the cognitive state transition of EEG data. Novel observations of individual neural markers showed that the beta power activity of the O1 and O2 sites contributed the most to the classifications, subjects exhibited decreased beta power in the ADHD group, and had larger decreases during cognitive execution. Significance: These findings showed that the proposed EEG-based LSTM networks are capable of extracting the varied temporal characteristics of high-resolution electrophysiological signals to differentiate between ADHD and NT children, and brought a new insight to facilitate the diagnosis of ADHD. The registration numbers of the institutional review boards are 16MMHIS021 and EC1070401-F.
Article
Full-text available
Recently, the use of portable electroencephalogram (EEG) devices to record brain signals in both health care monitoring and in other applications, such as fatigue detection in drivers, has been increased due to its low cost and ease of use. However, the measured EEG signals always mix with the electrooculogram (EOG), which are results due to eyelid blinking or eye movements. The eye-blinking/movement is an uncontrollable activity that results in a high-amplitude slow-time varying component that is mixed in the measured EEG signal. The presence of these artifacts misled our understanding of the underlying brain state. As the portable EEG devices comprise few EEG channels or sometimes a single EEG channel, classical artifact removal techniques such as blind source separation methods cannot be used to remove these artifacts from a single-channel EEG signal. Hence, there is a demand for the development of new single-channel-based artifact removal techniques. Singular spectrum analysis (SSA) has been widely used as a single-channel-based eye-blink artifact removal technique. However, while removing the artifact, the low-frequency components from the non-artifact region of the EEG signal are also removed by SSA. To preserve these low-frequency components, in this paper, we have proposed a new methodology by integrating the SSA with continuous wavelet transform (CWT) and the k-means clustering algorithm that removes the eye-blink artifact from the single-channel EEG signals without altering the low frequencies of the EEG signal. The proposed method is evaluated on both synthetic and real EEG signals. The results also show the superiority of the proposed method over the existing methods.
Article
Aim Information processing is supported by the cortico-cortical transmission of neural oscillations across brain regions. Recent studies have demonstrated that the rhythmic firing of neural populations is not random but is governed by interactions with other frequency bands. Specifically, the amplitude of gamma-band oscillations is associated with the phase of lower frequency oscillations in support of short and long-range communications among networks. This cross-frequency relation is thought to reflect the temporal coordination of neural communication. While schizophrenia patients show abnormal oscillatory responses across multiple frequencies at rest, it is unclear whether the functional relationships among frequency bands are intact. This study aimed to characterize the lower frequency (delta/theta, 1–8 Hz) phase and the amplitude of gamma oscillations in healthy subjects and schizophrenia patients at rest. Methods Low frequency-phase (delta- and theta- band) angles and gamma-band amplitude relationships were assessed in 142 schizophrenia patients and 128 healthy subjects. Results Significant low-frequency phase alteration related to high-power gamma was detected across broadly distributed scalp regions in both healthy subjects and patients. In patients, delta phase synchronization related to high-power gamma was significantly decreased at frontocentral, right middle temporal, and left temporoparietal electrodes but significantly increased at a left parietal electrode. Conclusions High-power gamma related delta phase alteration may reflect a core pathophysiologic abnormality in schizophrenia. Data-driven measures of functional relationships among frequency bands may prove useful in the development of novel therapeutics. Future studies are needed to determine whether these alterations are specific to schizophrenia or appear in other neuropsychiatric patient populations. This article is protected by copyright. All rights reserved.
Article
Full-text available
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.
Conference Paper
Full-text available
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.
Conference Paper
Full-text available
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.).
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Conference Paper
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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.
Article
Full-text available
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).
Article
Full-text available
The electroencephalogram (EEG) is often contaminated by electromyography (EMG). In this paper, a novel and robust technique is presented to eliminate EMG artifacts from EEG signals in real-time. First, the canonical correlation analysis (CCA) method is applied on the simulated EEG data contaminated by EMG and electrooculography (EOG) artifacts for separating EMG artifacts from EEG signals. The components responsible for EMG artifacts are distinguished from those responsible for brain activity based on the relative low autocorrelation. We demonstrate that the CCA method is more suitable to reconstruct the EMG-free EEG data than independent component analysis (ICA) methods. In addition, by applying CCA to analyze a number of EEG data contaminated by EMG artifacts, a correlation threshold is determined using an unbiased procedure. Hence, CCA can be used to remove EMG artifacts automatically. Finally, an example is given to verify that, after EMG artifacts were removed successfully from the EEG data contaminated by EMG and EOG simultaneously, not only the underlying brain activity signals but the EOG artifacts are preserved with little distortion.
Article
The electroencephalogram (EEG) provides a non-invasive, minimally restrictive, and relatively low-cost measure of mesoscale brain dynamics with high temporal resolution. Although signals recorded in parallel by multiple, near-adjacent EEG scalp electrode channels are highly-correlated and combine signals from many different sources, biological and non-biological, independent component analysis (ICA) has been shown to isolate the various source generator processes underlying those recordings. Independent components (IC) found by ICA decomposition can be manually inspected, selected, and interpreted, but doing so requires both time and practice as ICs have no order or intrinsic interpretations and therefore require further study of their properties. Alternatively, sufficiently-accurate automated IC classifiers can be used to classify ICs into broad source categories, speeding the analysis of EEG studies with many subjects and enabling the use of ICA decomposition in near-real-time applications. While many such classifiers have been proposed recently, this work presents the ICLabel project comprised of (1) the ICLabel dataset containing spatiotemporal measures for over 200,000 ICs from more than 6000 EEG recordings and matching component labels for over 6000 of those ICs, all using common average reference, (2) the ICLabel website for collecting crowdsourced IC labels and educating EEG researchers and practitioners about IC interpretation, and (3) the automated ICLabel classifier, freely available for MATLAB. The ICLabel classifier improves upon existing methods in two ways: by improving the accuracy of the computed label estimates and by enhancing its computational efficiency. The classifier outperforms or performs comparably to the previous best publicly available automated IC component classification method for all measured IC categories while computing those labels ten times faster than that classifier as shown by a systematic comparison against other publicly available EEG IC classifiers.