EEG Time Research Articles

Inter-subject transfer learning with an end-to-end deep convolutional neural network for EEG-based BCI

Objective. Despite the effective application of deep learning (DL) in brain–computer interface (BCI) systems, the successful execution of this technique, especially for inter-subject classification, in cognitive BCI has not been accomplished yet. In this paper, we propose a framework based on the deep convolutional neural network (CNN) to detect the attentive mental state from single-channel raw electroencephalography (EEG) data. Approach. We develop an end-to-end deep CNN to decode the attentional information from an EEG time series. We also explore the consequences of input representations on the performance of deep CNN by feeding three different EEG representations into the network. To ensure the practical application of the proposed framework and avoid time-consuming re-training, we perform inter-subject transfer learning techniques as a classification strategy. Eventually, to interpret the learned attentional patterns, we visualize and analyse the network perception of the attention and non-attention classes. Main results. The average classification accuracy is 79.26%, with only 15.83% of 120 subjects having an accuracy below 70% (a generally accepted threshold for BCI). This is while with the inter-subject approach, it is literally difficult to output high classification accuracy. This end-to-end classification framework surpasses conventional classification methods for attention detection. The visualization results demonstrate that the learned patterns from the raw data are meaningful. Significance. This framework significantly improves attention detection accuracy with inter-subject classification. Moreover, this study sheds light on the research on end-to-end learning; the proposed network is capable of learning from raw data with the least amount of pre-processing, which in turn eliminates the extensive computational load of time-consuming data preparation and feature extraction.

Brain stimulation method and analysis of eeg seizure time series

The aim of this research work is to develop keeping in mind the sweeping changes in Electro Convulsive Therapy (ECT). It is has been tried to inculcate these improvements to make the research work efficient and feasible. ECT is most effective treatment available to certain severe neurological disorders. Recent studies demonstrate association between certain parameters and methods of ECT seizure induction. There is evidence that the latter and therapeutic potency of ECT related. Seizure was obtained unilateral and bilateral application of electrical stimulus. An adequate seizure during ECT is still defined by its duration. Differing therapeutic potencies of ECT procedures cannot be explained on seizure duration alone; other parameter is spectral power in delta frequency band of EEG. This research work is to examine the clinical application of the findings to elucidate the therapeutic relevance of seizure measures other than seizure duration. The other measuring seizure parameters relevant therapeutic mechanism of ECT are power distribution, synchronization in various phases of EEG time series data collected during ECT and spectral power distribution in frequency bands of EEG. This research work involves several signal processing techniques: four channel EEG time series digital signal calibration, removing baseline drift, filtering processes to eliminate noise and artifacts, an automated selection of seizure phases. From the quantified EEG an empirical criterion for brain stimulus has been obtained.

Effect of linear mixing in EEG on synchronization and complex network measures studied using the Kuramoto model

Volume conduction in the brain may influence the synchronization between EEG signals considerably, as it may lead to detection of spurious functional couplings among the recording channels. It has been shown that the volume conduction effect can be approximated as a linear mixing of the electrical fields of the brain regions. In this paper, we investigate the reliability of various synchronization measures in the presence of the linear superposition in EEG time series. For this purpose, we applied linear mixing to artificially generated EEG times series using the Kuramoto model of coupled phase oscillators, which represents the behavior of coupled systems with local interactions at the fundamental level. Our simulation results showed that the phase-lag index and the synchronization measures based on the visibility graph algorithms were less sensitive to the linear mixing effect and could predict the coupling degree correctly even with strongly overlapping signals. The results of our further data analyses demonstrated the effect of linear superposition in time series on the behavior of various complex network measures. For each case, we provide recommendations for proper choices of synchronization measures to obtain complex network characteristics that are minimally sensitive to linear mixing.

BOLD-fMRI activity informed by network variation of scalp EEG in juvenile myoclonic epilepsy.

Epilepsy is marked by hypersynchronous bursts of neuronal activity, and seizures can propagate variably to any and all areas, leading to brain network dynamic organization. However, the relationship between the network characteristics of scalp EEG and blood oxygenation level-dependent (BOLD) responses in epilepsy patients is still not well known. In this study, simultaneous EEG and fMRI data were acquired in 18 juvenile myoclonic epilepsy (JME) patients. Then, the adapted directed transfer function (ADTF) values between EEG electrodes were calculated to define the time-varying network. The variation of network information flow within sliding windows was used as a temporal regressor in fMRI analysis to predict the BOLD response. To investigate the EEG-dependent functional coupling among the responding regions, modulatory interactions were analyzed for network variation of scalp EEG and BOLD time courses. The results showed that BOLD activations associated with high network variation were mainly located in the thalamus, cerebellum, precuneus, inferior temporal lobe and sensorimotor-related areas, including the middle cingulate cortex (MCC), supplemental motor area (SMA), and paracentral lobule. BOLD deactivations associated with medium network variation were found in the frontal, parietal, and occipital areas. In addition, modulatory interaction analysis demonstrated predominantly directional negative modulation effects among the thalamus, cerebellum, frontal and sensorimotor-related areas. This study described a novel method to link BOLD response with simultaneous functional network organization of scalp EEG. These findings suggested the validity of predicting epileptic activity using functional connectivity variation between electrodes. The functional coupling among the thalamus, frontal regions, cerebellum and sensorimotor-related regions may be characteristically involved in epilepsy generation and propagation, which provides new insight into the pathophysiological mechanisms and intervene targets for JME.

Chaotic System Design Based on Recurrent Artificial Neural Network for the Simulation of EEG Time Series

Electroencephalogram (EEG) signals captured from brain activities demonstrate chaotic features, and can be simulated by nonlinear dynamic time series outputs of chaotic systems. This article presents the research work of chaotic system generator design based on artificial neural network (ANN), for studying the chaotic features of human brain dynamics. The ANN training performances of Nonlinear Auto-Regressive (NAR) model are evaluated for the generation and prediction of chaotic system time series outputs, based on varying the ANN architecture and the precision of the generated training data. The NAR model is trained in open loop form with 1,000 training samples generated using Lorenz system equations and the forward Euler method. The close loop NAR model is used for the generation and prediction of Lorenz chaotic time series outputs. The training results show that better training performance can be achieved by increasing the number of feedback delays and the number of hidden neurons, at the cost of increasing the computational load.

Prediction of Electroencephalogram Time Series With Electro-Search Optimization Algorithm Trained Adaptive Neuro-Fuzzy Inference System

Nowadays, artificial intelligence is widely used in many biomedical-oriented problems. Because of obtained effective and efficient results, the use of intelligent solution mechanisms by artificial intelligence techniques is mainly focused on healthcare applications. Moving from the explanations, the objective of this paper is to provide an adaptive neuro-fuzzy inference system (ANFIS) trained by a recent optimization algorithm: electro-search optimization (ESO) for predicting the electroencephalogram (EEG) time series. In detail, the research was directed to the EEG time series showing chaotic characteristics so that an effective hybrid system can be designed for having an idea about future states of human brain activity in the case of possible diseases. The developed ANFIS-ESO system was evaluated with five different EEG time series and the obtained findings were reported accordingly. In addition, the ANFIS-ESO system was compared with alternative techniques-systems in order to see the performances according to different systems. In the end, it is possible to mention that the ANFIS-ESO system provides well-enough results in terms of predicting EEG time series. As a result of encouraging results, ANFIS-ESO is currently used actively for real cases.

Parameters of stochastic models for electroencephalogram data as biomarkers for child\u2019s neurodevelopment after cerebral malaria

The objective of this study was to test statistical features from the electroencephalogram (EEG) recordings as predictors of neurodevelopment and cognition of Ugandan children after coma due to cerebral malaria. The increments of the frequency bands of EEG time series were modeled as Student processes; the parameters of these Student processes were estimated and used along with clinical and demographic data in a machine-learning algorithm for the prediction of children’s neurodevelopmental and cognitive scores 6 months after cerebral malaria illness. The key innovation of this work is in the identification of stochastic EEG features that can serve as language-independent markers of the impact of cerebral malaria on the developing brain. The results can enhance prognostic determination of which children are in most need of rehabilitative interventions, which is especially important in resource-constrained settings such as sub-Saharan Africa.

Circadian Rhythms in Fractal Features of EEG Signals.

Time-of-day modulations affect both performance on a wide range of cognitive tasks and electrical activity of the brain, as recorded by electroencephalography (EEG). The aim of this work was to identify fluctuations of fractal properties of EEG time series due to circadian rhythms. In twenty-one healthy volunteers (all males, age between 20 and 30 years, chronotype: neutral type) high density EEG recordings at rest in open and closed eyes conditions were acquired in 4 times of the day (8.00 a.m., 11.30 a.m., 2.30 p.m., 7.00 p.m.). A vigilance task (Psychomotor Vigilance Test, PVT) was also performed. Detrended fluctuation Analysis (DFA) of envelope of alpha, beta and theta rhythms was performed, as well as Highuchi fractal dimension (HFD) of the whole band EEG. Our results evidenced circadian fluctuations of fractal features of EEG at rest in both eyes closed and eyes open conditions. Lower values of DFA exponent were found in the time T1 in closed eyes condition, likely effect of the sleep inertia. An alpha DFA exponent reduction was found also in central sensory-motor areas at time T3, the day time in which the sleepiness can be present. In eyes open condition, HFD lowered during the day. In eyes closed condition, an HFD increase was observed in central and frontal regions at time T2, the time in which alertness reaches its maximum and homeostatic sleep pressure is low. Complexity and the persistence of temporal correlations of brain rhythms change during daytime, parallel to changes in alertness and performance.

Constructing brain connectivity group graphs from EEG time series

ABSTRACTGraphical analysis of complex brain networks is a fundamental area of modern neuroscience. Functional connectivity is important since many neurological and psychiatric disorders, including schizophrenia, are described as ‘dys-connectivity’ syndromes. Using electroencephalogram time series collected on each of a group of 15 individuals with a common medical diagnosis of positive syndrome schizophrenia we seek to build a single, representative, brain functional connectivity group graph. Disparity/distance measures between spectral matrices are identified and used to define the normalized graph Laplacian enabling clustering of the spectral matrices for detecting ‘outlying’ individuals. Two such individuals are identified. For each remaining individual, we derive a test for each edge in the connectivity graph based on average estimated partial coherence over frequencies, and associated p-values are found. For each edge these are used in a multiple hypothesis test across individuals and the proportion rejecting the hypothesis of no edge is used to construct a connectivity group graph. This study provides a framework for integrating results on multiple individuals into a single overall connectivity structure.

A Multitaper Frequency-Domain Bootstrap Method.

Spectral properties of the electroencephalogram (EEG) are commonly analyzed to characterize the brain's oscillatory properties in basic science and clinical neuroscience studies. The spectrum is a function that describes power as a function of frequency. To date inference procedures for spectra have focused on constructing confidence intervals at single frequencies using large sample-based analytic procedures or jackknife techniques. These procedures perform well when the frequencies of interest are chosen before the analysis. When these frequencies are chosen after some of the data have been analyzed, the validity of these conditional inferences is not addressed. If power at more than one frequency is investigated, corrections for multiple comparisons must also be incorporated. To develop a statistical inference approach that considers the spectrum as a function defined across frequencies, we combine multitaper spectral methods with a frequency-domain bootstrap (FDB) procedure. The multitaper method is optimal for minimizing the bias-variance tradeoff in spectral estimation. The FDB makes it possible to conduct Monte Carlo based inferences for any part of the spectrum by drawing random samples that respect the dependence structure in the EEG time series. We show that our multitaper FDB procedure performs well in simulation studies and in analyses comparing EEG recordings of children from two different age groups receiving general anesthesia.

A hybrid unsupervised approach toward EEG epileptic spikes detection

Epileptic spikes are complementary sources of information in EEG to diagnose and localize the origin of epilepsy. However, not only is visual inspection of EEG labor intensive, time consuming, and prone to human error, but it also needs long-term training to acquire the level of skill required for identifying epileptic discharges. Therefore, computer-aided approaches were employed for the purpose of saving time and increasing the detection and source localization accuracy. One of the most important artifacts that may be confused as an epileptic spike, due to morphological resemblance, is eye blink. Only a few studies consider removal of this artifact prior to detection, and most of them used either visual inspection or computer-aided approaches, which need expert supervision. Consequently, in this paper, an unsupervised and EEG-based system with embedded eye blink artifact remover is developed to detect epileptic spikes. The proposed system includes three stages: eye blink artifact removal, feature extraction, and classification. Wavelet transform was employed for both artifact removal and feature extraction steps, and adaptive neuro-fuzzy inference system for classification purpose. The proposed method is verified using a publicly available EEG dataset. The results show the efficiency of this algorithm in detecting epileptic spikes using low-resolution EEG with least computational complexity, highest sensitivity, and lesser human interaction compared to similar studies. Moreover, since epileptic spike detection is a vital component of epilepsy source localization, therefore this algorithm can be utilized for EEG-based pre-surgical evaluation of epilepsy.

A research for single trial detection of error related negativity

Error related negativity (ERN) is generated in frontal and central cortical regions when individuals perceive errors. Because ERN has low signal-to-noise ratio and large individual difference, it is difficult for single trial ERN recognition. In current study, the optimized electroencephalograph (EEG) channels were selected based on the brain topography of ERN activity and ERN offline recognition rate, and the optimized EEG time segments were selected based on the ERN offline recognition rate, then the low frequency time domain and high frequency time-frequency domain features were analyzed based on wavelet transform, after which the ERN single detection algorithm was proposed based on the above procedures. Finally, we achieved average recognition rate of 72.0% ± 9.6% in 10 subjects by using the sample points feature in 0~3.9 Hz and the power and variance features in 3.9~15.6 Hz from the EEG segments of 200~600 ms on the selected 6 channels. Our work has the potential to help the error command real-time correction technique in the application of online brain-computer interface system.

Kurtosis-Based Detection of Intracranial High-Frequency Oscillations for the Identification of the Seizure Onset Zone

Pathological High-Frequency Oscillations (HFOs) have been recently proposed as potential biomarker of the seizure onset zone (SOZ) and have shown superior accuracy to interictal epileptiform discharges in delineating its anatomical boundaries. Characterization of HFOs is still in its infancy and this is reflected in the heterogeneity of analysis and reporting methods across studies and in clinical practice. The clinical approach to HFOs identification and quantification usually still relies on visual inspection of EEG data. In this study, we developed a pipeline for the detection and analysis of HFOs. This includes preliminary selection of the most informative channels exploiting statistical properties of the pre-ictal and ictal intracranial EEG (iEEG) time series based on spectral kurtosis, followed by wavelet-based characterization of the time-frequency properties of the signal. We performed a preliminary validation analyzing EEG data in the ripple frequency band (80-250 Hz) from six patients with drug-resistant epilepsy who underwent pre-surgical evaluation with stereo-EEG (SEEG) followed by surgical resection of pathologic brain areas, who had at least two-year positive post-surgical outcome. In this series, kurtosis-driven selection and wavelet-based detection of HFOs had average sensitivity of 81.94% and average specificity of 96.03% in identifying the HFO area which overlapped with the SOZ as defined by clinical presurgical workup. Furthermore, the kurtosis-based channel selection resulted in an average reduction in computational time of 66.60%.

Deep Classification of Epileptic Signals.

Electrophysiological observation plays a major role in epilepsy evaluation. However, human interpretation of brain signals is subjective and prone to misdiagnosis. Automating this process, especially seizure detection relying on scalpbased Electroencephalography (EEG) and intracranial EEG, has been the focus of research over recent decades. Nevertheless, its numerous challenges have inhibited a definitive solution. Inspired by recent advances in deep learning, here we describe a new classification approach for EEG time series based on Recurrent Neural Networks (RNNs) via the use of Long- Short Term Memory (LSTM) networks. The proposed deep network effectively learns and models discriminative temporal patterns from EEG sequential data. Especially, the features are automatically discovered from the raw EEG data without any pre-processing step, eliminating humans from laborious feature design task. Our light-weight system has a low computational complexity and reduced memory requirement for large training datasets. On a public dataset, a multi-fold cross-validation scheme of the proposed architecture exhibited an average validation accuracy of 95.54% and an average AUC of 0.9582 of the ROC curve among all sets defined in the experiment. This work reinforces the benefits of deep learning to be further attended in clinical applications and neuroscientific research.

A Distributed Descriptor Characterizing Structural Irregularity of EEG Time Series for Epileptic Seizure Detection.

This paper presents a novel descriptor aiming at anomaly detection in sequential data, like epileptic seizure detection with EEG time series. The descriptor is derived from the eigenvalue decomposition (EVD) of a Hankel-form data matrix generated from the raw time series. Simulation trials imply that the descriptor is capable of characterizing the structural aspect of a time series. In addition, we deploy the proposed descriptor as a feature extractor and apply it on Bonn Seizure Database which is widely used in seizure detection. The high accuracies on classification problems are comparable with the state-of-the-art so validate the effectiveness of our method.

Evaluating resective surgery targets in epilepsy patients: A comparison of quantitative EEG methods

BackgroundQuantitative analysis of intracranial EEG is a promising tool to assist clinicians in the planning of resective brain surgery in patients suffering from pharmacoresistant epilepsies. Quantifying the accuracy of such tools, however, is nontrivial as a ground truth to verify predictions about hypothetical resections is missing. New methodAs one possibility to address this, we use customized hypotheses tests to examine the agreement of the methods on a common set of patients. One method uses machine learning techniques to enable the predictive modeling of EEG time series. The other estimates nonlinear interrelation between EEG channels. Both methods were independently shown to distinguish patients with excellent post-surgical outcome (Engel class I) from those without improvement (Engel class IV) when assessing the electrodes associated with the tissue that was actually resected during brain surgery. Using the AND and OR conjunction of both methods we evaluate the performance gain that can be expected when combining them. ResultsBoth methods’ assessments correlate strongly positively with the similarity between a hypothetical resection and the corresponding actual resection in class I patients. Moreover, the Spearman rank correlation between the methods’ patient rankings is significantly positive. Comparison with existing method(s)To our best knowledge, this is the first study comparing surgery target assessments from fundamentally differing techniques. ConclusionsAlthough conceptually completely independent, there is a relation between the predictions obtained from both methods. Their broad consensus supports their application in clinical practice to provide physicians additional information in the process of presurgical evaluation.

T133. The effective brain connectome: An inverse problem

We give a general overview of recent developments in the study of Brain Connectivity. The essential task is to estimate the “Effective Brain Connectome” (EBC). This is the set of all direct interactions between neural masses. However, these interactions have to be estimated. We propose a formal definition of the EBC as a hypergraph defined on a multidimensional tensor. The simplest version of EBC is defined on a 3-dimensional tensor, with signatures for emitter regions, receiver regions and time delays, taking values for a Wiener-Akaike-Granger-Schweder influence measure. In the more complex case the dimensionality may take into consideration cortical layer, neurotransmitters etc. The support of the hypergraph is defined by the structural connectome. The actual challenge is to estimate the EBC from incomplete, noisy, multimodal neuroimaging data, especially from EEG and fMRI time series. Recognition of the mathematical structure of the underlying model allows the use of new tensor based methods for model inversion and data fusion. These are illustrated with resting state EEG/fMRI data. Finally, open problems in EBC studies are identifies and discussed.

F113. The methohexital challenge prior to intracranial endovascular embolization

Introduction The risks associated with endovascular embolization procedures are both mitigated and predicted with the use of pharmacologic provocative testing (PT) and intraoperative neurophysiologic monitoring (IONM). The utility of pharmacologic provocative testing has been reviewed in the literature, mostly in patients undergoing awake procedures with the use of neurologic examination and electroencephalogram. Here, we review our experience in patients undergoing intracranial endovascular embolization procedures with the use of IONM. Methods We reviewed our database to identify all patients undergoing endovascular procedures between January 1, 2014 and January 1, 2016. We included only patients who underwent cerebral endovascular embolization procedures with the use of methohexital PT. A retrospective chart review was performed to identify patients’ demographic factors, intraoperative findings, and postoperative examination details. We also reviewed perioperative notes to identify any adverse reactions to methohexital as well as other complications unrelated to the use of methohexital. Awake patients were tested with SSEP, EEG and real time neurologic examination while TcMEPs were performed in all anesthetized patients. BAEPs were performed in anesthetized patients if indicated. Methohexital was administered as an injection at a dose of 5 mg or 10 mg and repeat testing was performed if needed. Results A total of 64 endovascular procedures were performed in 31 patients with utilization of pharmacologic PT and were included in this review. PT was performed under general anesthesia in 54 procedures (84%) and awake testing was performed in 10 procedures (16%). PT was negative in 62 procedures and embolizations were performed in these patients. PT was positive in two procedures and the procedure was terminated without embolization in one patient. The other patient underwent embolization after repositioning the catheter at an alternative embolization site but repeat testing was not performed. There were no new postoperative neurologic deficits after any of these procedures. We calculated the specificity of pharmacologic PT as 100% as none of the patients with a negative provocative test developed a new postoperative neurologic deficit after embolization. Conclusion To the best of our knowledge, this is the first review on pharmacologic PT with methohexital and with the use of IONM under general anesthesia. Our results also indicate that IONM under general anesthesia can allow pharmacologic PT at a specificity comparable to awake testing. We also postulate that the absence of false negatives provides strong evidence that IONM accurately monitors the functional state of the brain during PT. If this were not the case, then new neurologic deficits should be seen in some of the cases with negative tests.

Functional brain connectivity analysis in amyotrophic lateral sclerosis: an EEG source-space study

Amyotrophic Lateral Sclerosis (ALS) is a severe neurodegenerative disease, which affects domains outside of the motor system. Recent findings have suggested that scalp electroencephalography (EEG) analysis may represent a useful tool for the monitoring of the disease. Here, we investigated functional connectivity in a set of ALS patients using source-reconstructed EEG time series comparing the results to age-matched healthy controls. Source based functional connectivity was estimated using a measure of amplitude envelope correlation (AEC) after performing orthogonalisation procedure in the time domain. The results of this study show that alpha band functional connectivity at source level discloses different patterns of inter-regional synchronization in ALS patients compared to healthy subjects. These findings suggest that ALS induces a strong frequency-specific reduction of functional connectivity patterns and that these changes are not limited to motor connections.

Deep Learning of EEG Time–Frequency Representations for Identifying Eye States

A new Convolutional Neural Network (CNN) architecture to classify nonstationary biomedical signals using their time–frequency representations is proposed. The present method uses the spectrogram of the biomedical signals as an input to CNN, in addition Non-negative matrix factorization (NMF) dictionary elements are used as an additional feature to improve the performance of the CNN model. Considering a number of applications involving eye state classification, such as in Parkinson’s disease detection, analysis of eye fatigue in 3D TVs, driver’s drowsiness detection, infant sleep-waking state identification, and classification of bipolar mood disorder and attention deficit hyperactivity, the proposed method was applied to Electroencephalography (EEG) data for classification of eye state. First, the spectrogram of EEG signal is obtained and used as an image input to CNN, simultaneously, the NMF feature is also fed to CNN. Further, both features are combined in fully connected layer of CNN architecture. The proposed method is compared with other existing methods for eye state detection and shows good classification accuracy with 96.16%. The prediction rate for the proposed method is 134 observations/second, which is suitable for brain–computer interface applications.