Raw EEG Research Articles

Classification of Electroencephalography Motor Execution Signals Using a Hybrid Neural Network Based on Instantaneous Frequency and Amplitude Obtained via Empirical Wavelet Transform

Brain–computer interfaces (BCIs) have garnered significant interest due to their potential to enable communication and control for individuals with limited or no ability to interact with technologies in a conventional way. By applying electrical signals generated by brain cells, BCIs eliminate the need for physical interaction with external devices. This study investigates the performance of traditional classifiers—specifically, linear discriminant analysis (LDA) and support vector machines (SVMs)—in comparison with a hybrid neural network model for EEG-based gesture classification. The dataset comprised EEG recordings of seven distinct gestures performed by 33 participants. Binary classification tasks were conducted using both raw windowed EEG signals and features extracted via bandpower and the empirical wavelet transform (EWT). The hybrid neural network architecture demonstrated higher classification accuracy compared to the standard classifiers. These findings suggest that combining featuring extraction with deep learning models offers a promising approach for improving EEG gesture recognition in BCI systems.

EEG-Based Seizure Detection Using Dual-Branch CNN-ViT Network Integrating Phase and Power Spectrograms.

Background/Objectives: Epilepsy is a common neurological disorder with pathological mechanisms closely associated with the spatiotemporal dynamic characteristics of electroencephalogram (EEG) signals. Although significant progress has been made in epileptic seizure detection methods using time-frequency analysis, current research still faces challenges in terms of an insufficient utilization of phase information. Methods: In this study, we propose an effective epileptic seizure detection framework based on continuous wavelet transform (CWT) and a hybrid network consisting of convolutional neural network (CNN) and vision transformer (ViT). First, the raw EEG signals are processed by the CWT. Then, the phase spectrogram and power spectrogram of the EEG are generated, and they are sent into the designed CNN and ViT branches of the network to extract more discriminative EEG features. Finally, the features output from the two branches are fused and fed into the classification network to obtain the detection results. Results: Experimental results on the CHB-MIT public dataset and our SH-SDU clinical dataset show that the proposed framework achieves sensitivities of 98.09% and 89.02%, specificities of 98.21% and 95.46%, and average accuracies of 98.45% and 94.66%, respectively. Furthermore, we compared the spectral characteristics of CWT with other time-frequency transforms within the hybrid architecture, demonstrating the advantages of the CWT-based CNN-ViT architecture. Conclusions: These results highlight the outstanding epileptic seizure detection performance of the proposed framework and its significant clinical feasibility.

Universal semantic feature extraction from EEG signals: a task-independent framework

Objective.Extracting universal, task-independent semantic features from electroencephalography (EEG) signals remains an open challenge. Traditional approaches are often task-specific, limiting their generalization across different EEG paradigms. This study aims to develop a robust, unsupervised framework for learning high-level, task-independent neural representations.Approach.We propose a novel framework integrating convolutional neural networks, AutoEncoders, and Transformers to extract both low-level spatiotemporal patterns and high-level semantic features from EEG signals. The model is trained in an unsupervised manner to ensure adaptability across diverse EEG paradigms, including motor imagery (MI), steady-state visually evoked potentials (SSVEPs), and event-related potentials (ERPs, specifically P300). Extensive analyses, including clustering, correlation, and ablation studies, are conducted to validate the quality and interpretability of the extracted features.Main results.Our method achieves state-of-the-art performance, with average classification accuracies of 83.50% and 84.84% on MI datasets (BCICIV_2a and BCICIV_2b), 98.41% and 99.66% on SSVEP datasets (Lee2019-SSVEP and Nakanishi2015), and an average AUC of 91.80% across eight ERP datasets. t-distributed stochastic neighbor embedding and clustering analyses reveal that the extracted features exhibit enhanced separability and structure compared to raw EEG data. Correlation studies confirm the framework's ability to balance universal and subject-specific features, while ablation results highlight the near-optimality of the selected model configuration.Significance.This work establishes a universal framework for task-independent semantic feature extraction from EEG signals, bridging the gap between conventional feature engineering and modern deep learning methods. By providing robust, generalizable representations across diverse EEG paradigms, this approach lays the foundation for advanced brain-computer interface applications, cross-task EEG analysis, and future developments in semantic EEG processing.

Music interventions and obstructive sleep apnea: a brain connectivity analysis

BackgroundThe blockage in the upper airway that occurs, while sleeping is represented as obstructive sleep apnea (OSA). This seem to be a major issue which cause breathing difficulties also increases the risk of severe complications, such as heart attacks and strokes. Therefore, in this proposed study the impact of OSA using brain connectivity analysis under various conditions such as Neelambari, Kapi, and no music has been investigated. The electroencephalogram (EEG) recordings of twelve subjects were acquired in two different conditions, such as listening to music 1 and 2 (Neelambari and Kapi) and the absence of music. The raw EEG signals were then pre-processed using both bandpass and notch filters. Meanwhile, the EEG sub-bands were obtained using the wavelet packet decomposition (WPD) method. These sub-bands, including delta, theta, alpha, and beta, were used for brain connectivity analysis. This approach provides the visualization of frequency-specific regional brain connectivity patterns by applying Pearson Correlation to the absolute values of the detail coefficients from WPD using a graph theory metric, node strength.ResultsIncreased connectivity in the right hemisphere of the brain was observed among the nodes in the frontal and temporal regions (F8, FC6, and T8) when participants listened to Neelambari music (Music 1). In the beta band, the correlation values for Neelambari music ranged from a minimum of 0.943 to a maximum of 0.998. In the delta band, positive correlation values ranged from 0.945 (minimum) to 0.999 (maximum). The alpha and theta bands exhibited moderate correlations, ranging from 0.746 (minimum) to 0.996 (maximum). Compared to Kapi music, Neelambari music showed stronger neural synchronization, evidenced by consistently higher correlation values across all frequency bands. This increased connectivity suggests that Neelambari music may profoundly impact brain dynamics, potentially enhancing cognitive or physiological responses.ConclusionsIn conclusion, it has been analyzed that OSA patients have positive brain connectivity while listening to music 1 (Neelambari).

Electroencephalogram-Based Anesthesia Indices Differently React to Modulations of Alpha-Oscillatory Activity.

The electroencephalographic (EEG) provides the anesthesiologist with information regarding the level of anesthesia. Processed EEG indices are available that reflect the level of anesthesia as a single number. Strong oscillatory EEG activity in the alpha-band may be associated with an adequate level of anesthesia and a lower incidence of cognitive sequelae. So far, we do not know how the processed indices would react to changes in the alpha-band activity. Hence, we modulated the alpha-oscillatory activity of intraoperative EEG to assess possible index changes. We performed our analyses based on data from 2 studies. Intraoperative EEG was extracted, and we isolated the alpha-band activity by band-pass filtering (8-12 Hz). We added or subtracted this activity to the original EEG in different steps with different amplifications of the alpha signal. We then replayed these signals to the bispectral index (BIS), the Entropy Module (state entropy [SE]), the CONOX (qCON), and the SEDLine (patient state index [PSI]); and evaluated the alpha-band modulation's impact on the respective index. The indices behaved differently to the modulation. In general, indices decreased with stronger alpha-band activity, but the rate of change was different with SE showing the strongest change (9% per step) and PSI and BIS ( < 5% per step) showing the weakest change. A simple regression analysis revealed a decrease of 0.02 to 0.09 index points with increasing alpha amplification. While the alpha-band in the intraoperative EEG seems to carry information regarding the quality of anesthesia, changes in the alpha-band activity do neither strongly nor uniformly influence processed EEG indices. Hence, to assess alpha-oscillatory activity's strength, the user needs to focus on the raw EEG or its spectral representation also displayed on the monitoring systems.

Cortical, subcortical, brainstem and autonomic responses to nociception under total intravenous anesthesia.

Cortical, subcortical, brainstem and autonomic responses to nociception under total intravenous anesthesia.

Dataset of 37-subject EEG recordings using a low-cost mobile EEG headset during a semantic relatedness judgment task.

Dataset of 37-subject EEG recordings using a low-cost mobile EEG headset during a semantic relatedness judgment task.

A Fine-grained Hemispheric Asymmetry Network for accurate and interpretable EEG-based emotion classification.

A Fine-grained Hemispheric Asymmetry Network for accurate and interpretable EEG-based emotion classification.

Enhancing Schizophrenia Diagnosis Through Multi-View EEG Analysis: Integrating Raw Signals and Spectrograms in a Deep Learning Framework.

Objective: Schizophrenia is a chronic mental disorder marked by symptoms such as hallucinations, delusions, and cognitive impairments, which profoundly affect individuals' lives. Early detection is crucial for improving treatment outcomes, but the diagnostic process remains complex due to the disorder's multifaceted nature. In recent years, EEG data have been increasingly investigated to detect neural patterns linked to schizophrenia. Methods: This study presents a deep learning framework that integrates both raw multi-channel EEG signals and their spectrograms. Our two-branch model processes these complementary data views to capture both temporal dynamics and frequency-specific features while employing depth-wise convolution to efficiently combine spatial dependencies across EEG channels. Results: The model was evaluated on two datasets, consisting of 84 and 28 subjects, achieving classification accuracies of 0.985 and 0.994, respectively. These results highlight the effectiveness of combining raw EEG signals with their time-frequency representations for precise and automated schizophrenia detection. Additionally, an ablation study assessed the contributions of different architectural components. Conclusions: The approach outperformed existing methods in the literature, underscoring the value of utilizing multi-view EEG data in schizophrenia detection. These promising results suggest that our framework could contribute to more effective diagnostic tools in clinical practice.

MSBiLSTM-Attention: EEG Emotion Recognition Model Based on Spatiotemporal Feature Fusion.

Emotional states play a crucial role in shaping decision-making and social interactions, with sentiment analysis becoming an essential technology in human-computer emotional engagement, garnering increasing interest in artificial intelligence research. In EEG-based emotion analysis, the main challenges are feature extraction and classifier design, making the extraction of spatiotemporal information from EEG signals vital for effective emotion classification. Current methods largely depend on machine learning with manual feature extraction, while deep learning offers the advantage of automatic feature extraction and classification. Nonetheless, many deep learning approaches still necessitate manual preprocessing, which hampers accuracy and convenience. This paper introduces a novel deep learning technique that integrates multi-scale convolution and bidirectional long short-term memory networks with an attention mechanism for automatic EEG feature extraction and classification. By using raw EEG data, the method applies multi-scale convolutional neural networks and bidirectional long short-term memory networks to extract and merge features, selects key features via an attention mechanism, and classifies emotional EEG signals through a fully connected layer. The proposed model was evaluated on the SEED dataset for emotion classification. Experimental results demonstrate that this method effectively classifies EEG-based emotions, achieving classification accuracies of 99.44% for the three-class task and 99.85% for the four-class task in single validation, with average 10-fold-cross-validation accuracies of 99.49% and 99.70%, respectively. These findings suggest that the MSBiLSTM-Attention model is a powerful approach for emotion recognition.

Identification of Alzheimer’s disease brain networks based on EEG phase synchronization

ObjectiveUsing the phase synchronization of EEG signals, two different phases, PLI and PLV, were used to construct brain network analysis and graph convolutional neural network, respectively, to achieve automatic identification of Alzheimer’s disease (AD) and to assist in the early diagnosis of Alzheimer’s disease.MethodsIn this paper, we selected outpatients (16 AD subjects, 20 mild cognitive impairment (MCI) subjects and 21 healthy control (HC) subjects) from the outpatient clinic of Yangpu Mental Health Center in Shanghai, China, from January 2023 to December 2023, and collected resting-state EEG data. To collect resting-state EEG data, each patient was asked to sit down with eyes closed for 5 min. Firstly, the acquired EEG data were preprocessed to extract the data in the α-band at 8–13 Hz; secondly, the phase lag index (PLI) and phase-locked value (PLV) were used to construct the brain functional network, and the brain functional connectivity map was visualized by brain functional connectivity analysis. Finally, the constructed PLI and PLV were input into the graph convolutional neural network (GCN) model as node features for training and classification, respectively.ResultsHealthy controls had relatively strong mean brain functional connectivity in the PLV brain network compared to AD and MCI patients. MCI patients showed lower mean brain functional connectivity in the brain network of PLI, while all three groups showed significant differences in brain functional connectivity between parietal and occipital lobes. The GCN model improved classification accuracy by more than 10% compared to using a machine learning classifier. When PLV was used as the nodal feature in the GCN model, the model achieved an average classification accuracy of 77.80% for the three groups of AD, MCI and HC, which was an improvement over the accuracy of choosing raw EEG data and PLI as the nodal feature. The performance of the model was further validated.ConclusionsThe experimental results show that the GCN model can effectively identify the graph structure compared with the traditional machine learning model, the GCN-PLV model can better classify AD patients, and the alpha band is proved to be more suitable for AD resting-state EEG by tenfold cross-validation. The brain network map constructed based on PLI and PLV can further capture the local features of EEG signals and the intrinsic functional relationships between brain regions, and the combination of these two models has certain reference value for the diagnosis of AD patients.

EEG-based recognition of hand movement and its parameter

Objecitve. Brain-computer interface is a cutting-edge technology that enables interaction with external devices by decoding human intentions, and is highly valuable in the fields of medical rehabilitation and human-robot collaboration. The technique of decoding motor intent for motor execution (ME) based on electroencephalographic (EEG) signals is in the feasibility study stage by now. There are still insufficient studies on the accuracy of ME EEG signal recognition in between-subjects classification to reach the level of realistic applications. This paper aims to investigate EEG signal-based hand movement recognition by analyzing low-frequency time-domain information.Approach. Experiments with four types of hand movements, two force parameter (picking up and pushing) tasks, and a four-target directional displacement task were designed and executed, and the EEG data from thirteen healthy volunteers was collected. Sliding window approach is used to expand the dataset in order to address the issue of EEG signal overfitting. Furtherly, Convolutional Neural Network (CNN)-Bidirectional Long Short-Term Memory Network (BiLSTM) model, an end-to-end serial combination of a BiLSTM and (CNN) is constructed to classify and recognize the hand movement based on the raw EEG data.Main results. According to the experimental results, the model is able to categorize four types of hand movements, picking up movements, pushing movements, and four target direction displacement movements with an accuracy of 99.14% ± 0.49%, 99.29% ± 0.11%, 99.23% ± 0.60%, and 98.11% ± 0.23%, respectively.Significance. Furthermore, comparative tests conducted with alternative deep learning models (LSTM, CNN, EEGNet, CNN-LSTM) demonstrates that the CNN-BiLSTM model is with practicable accuracy in terms of EEG-based hand movement recognition and its parameter decoding.

Deep Riemannian Networks for End-to-End EEG Decoding

Abstract State-of-the-art performance in electroencephalography (EEG) decoding tasks is currently often achieved with either Deep-Learning (DL) or Riemannian-Geometry-based decoders (RBDs). Recently, there is growing interest in Deep Riemannian Networks (DRNs) possibly combining the advantages of both previous classes of methods. However, there are still a range of topics where additional insight is needed to pave the way for a more widespread application of DRNs in EEG. These include architecture design questions such as network size and end-to-end ability. How these factors affect model performance has not been explored. Additionally, it is not clear how the data within these networks is transformed, and whether this would correlate with traditional EEG decoding. Our study aims to lay the groundwork in the area of these topics through the analysis of DRNs for EEG with a wide range of hyperparameters. Networks were tested on five public EEG datasets and compared with state-of-the-art ConvNets. Here we propose end-to-end EEG SPDNet (EE(G)-SPDNet), and we show that this wide, end-to-end DRN can outperform the ConvNets, and in doing so use physiologically plausible frequency regions. We also show that the end-to-end approach learns more complex filters than traditional band-pass filters targeting the classical alpha, beta, and gamma frequency bands of the EEG, and that performance can benefit from channel specific filtering approaches. Additionally, architectural analysis revealed areas for further improvement due to the possible under utilisation of Riemannian specific information throughout the network. Our study thus shows how to design and train DRNs to infer task-related information from the raw EEG without the need of handcrafted filterbanks and highlights the potential of end-to-end DRNs such as EE(G)-SPDNet for high-performance EEG decoding.

Cognitive load recognition in simulated flight missions: an EEG study.

Cognitive load recognition (CLR) utilizing EEG signals has experienced significant advancement in recent years. However, current load-eliciting paradigms often rely on simplistic cognitive tasks such as arithmetic calculations, failing to adequately replicate real-world scenarios and lacking applicability. This study explores simulated flight missions over time to better reflect operational environments and investigate temporal dynamics of multiple load states. Thirty-six participants were recruited to perform simulated flight tasks with varying cognitive load levels of low, medium, and high. Throughout the experiments, we collected EEG load data from three sessions, pre- and post-task resting-state EEG data, subjective ratings, and objective performance metrics. Then, we employed several deep convolutional neural network (CNN) models, utilizing raw EEG data as model input, to assess cognitive load levels with six classification designs. Key findings from the study include (1) a notable distinction between resting-state and post-fatigue EEG data; (2) superior performance of shallow CNN models compared to more complex ones; and (3) temporal dynamics decline in CLR as the missions progressed. This paper establishes a potential foundation for assessing cognitive states during intricate simulated tasks across different individuals.

Effects of bupivacaine and morphine on the electroencephalogram and postoperative pain in castrated dogs.

Effects of bupivacaine and morphine on the electroencephalogram and postoperative pain in castrated dogs.

Electroencephalography Decoding with Conditional Identification Generator.

Decoding Electroencephalography (EEG) signals are extremely useful for advancing and understanding human-artificial intelligence (AI) interaction systems. Recent advancements in deep neural networks (DNNs) have demonstrated significant promise in this respect due to their ability to model complex nonlinear relationships. However, DNNs face persistent challenges in addressing the inter-person variability inherent in EEG signals, which limits their generalizability. To tackle this limitation, we propose a novel framework that integrates conditional identification information, leveraging the interaction between EEG signals and individual traits to enhance the model's internal representation and improve decoding accuracy. Building on this foundation, we further introduce a privacy-preserving conditional information generator - a generative model that derives embedding knowledge directly from raw EEG signals. This approach eliminates the need for personal identification via individual tests, ensuring both efficiency and privacy. Experimental evaluations conducted on WithMe dataset confirm that this framework outperforms baseline network architectures. Notably, our approach achieves substantial improvements in decoding accuracy for both familiar and unseen subjects, paving the way for efficient, robust, and privacy-conscious human-computer interface systems.

Early attention-deficit/hyperactivity disorder (ADHD) with NeuroDCT-ICA and rhinofish optimization (RFO) algorithm based optimized ADHD-AttentionNet

The ADHD detector analyzes behavioral, cognitive, or physiological data (e.g., EEG, eye-tracking, or surveys) to identify patterns associated with ADHD symptoms. This work offers a more sophisticated method of detecting ADHD by overcoming the main drawbacks of existing approaches in terms of data processing, detection accuracy, and computational time. The work is inspired by the fact that Deep Learning (DL) frameworks could transform the existing detection systems of ADHD. In the proposed framework, there is a new NeuroDCT-ICA module for the preprocessing of raw EEG data, which guarantees the elimination of noise and extraction of informative features. Moreover, the method introduces a novel RhinoFish Optimization (RFO) algorithm for selecting optimal features, which enhance the data processing capacity and the stability of the system. As a core of the approach, there is the ADHD-AttentionNet – the deep learning-based model aimed at improving the accuracy and confidence of ADHD identification. The model is validated with the standard metrics, and the performance of the model is outstanding as it has high accuracy of 98.52%, F-score of 98.26% and specificity of 98.16%. These outcomes show that the proposed model yields better accuracy in detecting ADHD related patterns.

Hybrid CNN-GRU Models for Improved EEG Motor Imagery Classification.

Brain-computer interfaces (BCIs) based on electroencephalography (EEG) enable neural activity interpretation for device control, with motor imagery (MI) serving as a key paradigm for decoding imagined movements. Efficient feature extraction from raw EEG signals is essential to improve classification accuracy while minimizing reliance on extensive preprocessing. In this study, we introduce new hybrid architectures to enhance MI classification using data augmentation and a limited number of EEG channels. The first model combines a shallow convolutional neural network and a gated recurrent unit (CNN-GRU), while the second incorporates a convolutional neural network with a bidirectional gated recurrent unit (CNN-Bi-GRU). Evaluated using the publicly available PhysioNet dataset, the CNN-GRU classifier achieved peak mean accuracy rates of 99.71%, 99.73%, 99.61%, and 99.86% for tasks involving left fist (LF), right fist (RF), both fists (LRF), and both feet (BF), respectively. The experimental results provide compelling evidence that our proposed models outperform current state-of-the-art methods, underscoring their efficiency on small-scale EEG datasets. The CNN-GRU and CNN-Bi-GRU architectures exhibit superior predictive reliability, offering a faster, cost-effective solution for user-adaptable MI-BCI applications.

Lower limb motor imagery EEG dataset based on the multi-paradigm and longitudinal-training of stroke patients

Motor dysfunction is one of the most significant sequelae of stroke, with lower limb impairment being a major concern for stroke patients. Motor imagery (MI) technology based on brain-computer interface (BCI) offers promising rehabilitation potential for stroke patients by activating motor-related brain areas. However, developing a robust BCI-MI system and uncovering the underlying mechanisms of neural plasticity during stroke recovery through such systems requires large-scale datasets. These datasets are particularly needed for accurate lower limb MI in stroke patients and for longitudinal data reflecting the rehabilitation process. This study addresses this gap by collecting EEG data from 27 stroke patients, covering two enhanced paradigms and three different time points. The dataset includes raw EEG signals, preprocessed data, and patient information. An initial analysis using CSP-SVM on the dataset yielded an average classification accuracy of 80.50%. We anticipate that this dataset will facilitate research into brain neuroplasticity in stroke patients, aid in the development of decoding algorithms for lower limb stroke, and contribute to the establishment of comprehensive stroke rehabilitation systems.

A Comprehensive Review of Modern EEG Signal Processing for Emotional and Stress Detection: From Data to Insight

Using the brain's thoughts to communicate with the outside world is known as brain-computer interface. Electroencephalography (EEG) has emerged as a powerful non-invasive method for understanding brain activity and developing applications in neurology, medicine, and human-computer interaction. Robust signal processing techniques, which convert raw EEG data into meaningful insights, are crucial to the effectiveness of EEG-based systems. This paper aimed to discuss the various approaches that need to be adjusted for each stage of brain signal processing, with a number of approaches and experimental results reported and compared in previous surveys. The advantages and disadvantages of each method, such as signal capture, signal augmentation, feature extraction, and signal classification, are easily and clearly interpreted in this work.