Deep Representation-Based Domain Adaptation for Nonstationary EEG Classification

Cross-subject mental workload recognition using bi-classifier domain adversarial learning

To deploy Electroencephalogram (EEG) based Mental Workload Recognition (MWR) systems in the real world, it is crucial to develop general models that can be applied across subjects. Previous studies have utilized domain adaptation to mitigate inter-subject discrepancies in EEG data distributions. However, they have focused on reducing global domain discrepancy, while neglecting local workload-categorical domain divergence. This degrades the workload-discriminating ability of subject-invariant features. To deal with this problem, we propose a novel joint category-wise and domain-wise alignment Domain Adaptation (cdaDA) algorithm, using bi-classifier learning and domain discriminative adversarial learning. The bi-classifier learning approach is adopted to address the similarities and differences between categories, helping to align EEG data within the same mental workload categories. Additionally, the domain discriminative adversarial learning technique is adopted to consider global domain information by minimizing global domain discrepancy. By integrating both local category information and global domain information, the cdaDA model performs a coarse-to-fine alignment and achieves promising cross-subject MWR results.

HATNet: EEG-Based Hybrid Attention Transfer Learning Network for Train Driver State Detection

Electroencephalography (EEG) is widely utilized for train driver state detection due to its high accuracy and low latency. However, existing methods for driver status detection rarely use the rich physiological information in EEG to improve detection performance. Moreover, there is currently a lack of EEG datasets for abnormal states of train drivers. To address these gaps, we propose a novel transfer learning model based on a hybrid attention mechanism, named hybrid attention-based transfer learning network (HATNet). We first segment the EEG signals into patches and utilize the hybrid attention module to capture local and global temporal patterns. Then, a channel-wise attention module is introduced to establish spatial representations among EEG channels. Finally, during the training process, we employ a calibration-based transfer learning strategy, which allows for adaptation to the EEG data distribution of new subjects using minimal data. To validate the effectiveness of our proposed model, we conduct a multistimulus oddball experiment to establish a EEG dataset of abnormal states for train drivers. Experimental results on this dataset indicate that: 1) Compared to the state-of-the-art end-to-end models, HATNet achieves the highest classification accuracy in both subject-dependent and subject-independent tasks at 94.26% and 87.03%, respectively, and 2) The proposed hybrid attention module effectively captures the temporal semantic information of EEG data.

EEG-DG: A Multi-Source Domain Generalization Framework for Motor Imagery EEG Classification

Motorimagery EEG classification plays a crucial role in non-invasive Brain-Computer Interface (BCI) research. However, the performance of classification is affected by the non-stationarity and individual variations of EEG signals. Simply pooling EEG data with different statistical distributions to train a classification model can severely degrade the generalization performance. To address this issue, the existing methods primarily focus on domain adaptation, which requires access to the test data during training. This is unrealistic and impractical in many EEG application scenarios. In this paper, we propose a novel multi-source domain generalization framework called EEG-DG, which leverages multiple source domains with different statistical distributions to build generalizable models on unseen target EEG data. We optimize both the marginal and conditional distributions to ensure the stability of the joint distribution across source domains and extend it to a multi-source domain generalization framework to achieve domain-invariant feature representation, thereby alleviating calibration efforts. Systematic experiments conducted on a simulative dataset, BCI competition IV 2a, 2b, and OpenBMI datasets, demonstrate the superiority and competitive performance of our proposed framework over other state-of-the-art methods. Specifically, EEG-DG achieves average classification accuracies of 81.79% and 87.12% on datasets IV-2a and IV-2b, respectively, and 78.37% and 76.94% for inter-session and inter-subject evaluations on dataset OpenBMI, which even outperforms some domain adaptation methods.

Feature-aware domain invariant representation learning for EEG motor imagery decoding

Electroencephalography (EEG)-based motor imagery (MI) is extensively utilized in clinical rehabilitation and virtual reality-based movement control. Decoding EEG-based MI signals is challenging because of the inherent spatio-temporal variability of the original signal representation, coupled with a low signal-to-noise ratio (SNR), which impedes the extraction of clean and robust features. To address this issue, we propose a multi-scale spatio-temporal domain-invariant representation learning method, termed MSDI. By decomposing the original signal into spatial and temporal components, the proposed method extracts invariant features at multiple scales from both components. To further constrain the representation to invariant domains, we introduce a feature-aware shift operation that resamples the representation based on its feature statistics and feature measure, thereby projecting the features into a domain-invariant space. We evaluate our proposed method via two publicly available datasets, BNCI2014-001 and BNCI2014-004, demonstrating state-of-the-art performance on both datasets. Furthermore, our method exhibits superior time efficiency and noise resistance.

Subject-Adaptation Salient Wave Detection Network for Multimodal Sleep Stage Classification

Sleep stage classification is an important step in the diagnosis and treatment of sleep disorders. Despite the high classification performance of previous sleep stage classification work, some challenges remain unresolved: 1) How to effectively capture salient waves in sleep signals to improve sleep stage classification results. 2) How to capture salient waves affected by inter-subject variability. 3) How to adaptively regulate the importance of different modals for different sleep stages. To address these challenges, we propose SleepWaveNet, a multimodal salient wave detection network, which is motivated by the salient object detection task in computer vision. It has a U-Transformer structure to detect salient waves in sleep signals. Meanwhile, the subject-adaptation wave extraction architecture based on transfer learning can adapt to the information of target individuals and extract salient waves with inter-subject variability. In addition, the multimodal attention module can adaptively enhance the importance of specific modal data for sleep stage classification tasks. Experiments on three datasets show that SleepWaveNet has better overall performance than existing baselines. Moreover, visualization experiments show that the model has the ability to capture salient waves with inter-subject variability.

Deep Multiview Module Adaption Transfer Network for Subject-Specific EEG Recognition

Transfer learning is one of the popular methods to solve the problem of insufficient data in subject-specific electroencephalogram (EEG) recognition tasks. However, most existing approaches ignore the difference between subjects and transfer the same feature representations from source domain to different target domains, resulting in poor transfer performance. To address this issue, we propose a novel subject-specific EEG recognition method named deep multiview module adaption transfer (DMV-MAT) network. First, we design a universal deep multiview (DMV) network to generate different types of discriminative features from multiple perspectives, which improves the generalization performance by extensive feature sets. Second, module adaption transfer (MAT) is designed to evaluate each module by the feature distributions of source and target samples, which can generate an optimal weight sharing strategy for each target subject and promote the model to learn domain-invariant and domain-specific features simultaneously. We conduct extensive experiments in two EEG recognition tasks, i.e., motor imagery (MI) and seizure prediction, on four datasets. Experimental results demonstrate that the proposed method achieves promising performance compared with the state-of-the-art methods, indicating a feasible solution for subject-specific EEG recognition tasks. Implementation codes are available at https://github.com/YangLibuaa/DMV-MAT.

MI-Mamba: A hybrid motor imagery electroencephalograph classification model with Mamba's global scanning

Deep learning has revolutionized electroencephalograph (EEG) decoding, with convolutional neural networks (CNNs) being a predominant tool. However, CNNs struggle with long-term dependencies in sequential EEG data. Models like long short-term memory and transformers improve performance but still face challenges of computational efficiency and long sequences. Mamba, a state space model-based method, excels in modeling long sequences. To overcome the limitations of existing EEG decoding models and exploit Mamba's potential in EEG analysis, we propose MI-Mamba, a model integrating CNN with Mamba for motor imagery (MI) data decoding. MI-Mamba processes multi-channel EEG signals through a single convolutional layer to capture spatial features in the local temporal domain, followed by a Mamba module that processes global temporal features. A fully connected, layer-based classifier is used to derive classification results. Evaluated on two public MI datasets, MI-Mamba achieves 80.59% accuracy in the four-class MI task of the BCI Competition IV 2a dataset and 84.42% in the two-class task of the BCI Competition IV 2b dataset, while reducing parameter count by nearly six times compared to the most advanced previous models. These results highlight MI-Mamba's effectiveness in MI decoding and its potential as a new backbone for general EEG decoding.

Multiscale Spatial-Temporal Feature Fusion Neural Network for Motor Imagery Brain-Computer Interfaces

Motor imagery, one of the main brain-computer interface (BCI) paradigms, has been extensively utilized in numerous BCI applications, such as the interaction between disabled people and external devices. Precise decoding, one of the most significant aspects of realizing efficient and stable interaction, has received a great deal of intensive research. However, the current decoding methods based on deep learning are still dominated by single-scale serial convolution, which leads to insufficient extraction of abundant information from motor imagery signals. To overcome such challenges, we propose a new end-to-end convolutional neural network based on multiscale spatial-temporal feature fusion (MSTFNet) for EEG classification of motor imagery. The architecture of MSTFNet consists of four distinct modules: feature enhancement module, multiscale temporal feature extraction module, spatial feature extraction module and feature fusion module, with the latter being further divided into the depthwise separable convolution block and efficient channel attention block. Moreover, we implement a straightforward yet potent data augmentation strategy to bolster the performance of MSTFNet significantly. To validate the performance of MSTFNet, we conduct cross-session experiments and leave-one-subject-out experiments. The cross-session experiment is conducted across two public datasets and one laboratory dataset. On the public datasets of BCI Competition IV 2a and BCI Competition IV 2b, MSTFNet achieves classification accuracies of 83.62% and 89.26%, respectively. On the laboratory dataset, MSTFNet achieves 86.68% classification accuracy. Besides, the leave-one-subject-out experiment is performed on the BCI Competition IV 2a dataset, and MSTFNet achieves 66.31% classification accuracy. These experimental results outperform several state-of-the-art methodologies, indicate the proposed MSTFNet's robust capability in decoding EEG signals associated with motor imagery.

Adaptive deep feature representation learning for cross-subject EEG decoding

BACKGROUND

The collection of substantial amounts of electroencephalogram (EEG) data is typically time-consuming and labor-intensive, which adversely impacts the development of decoding models with strong generalizability, particularly when the available data is limited. Utilizing sufficient EEG data from other subjects to aid in modeling the target subject presents a potential solution, commonly referred to as domain adaptation. Most current domain adaptation techniques for EEG decoding primarily focus on learning shared feature representations through domain alignment strategies. Since the domain shift cannot be completely removed, target EEG samples located near the edge of clusters are also susceptible to misclassification.

METHODS

We propose a novel adaptive deep feature representation (ADFR) framework to improve the cross-subject EEG classification performance through learning transferable EEG feature representations. Specifically, we first minimize the distribution discrepancy between the source and target domains by employing maximum mean discrepancy (MMD) regularization, which aids in learning the shared feature representations. We then utilize the instance-based discriminative feature learning (IDFL) regularization to make the learned feature representations more discriminative. Finally, the entropy minimization (EM) regularization is further integrated to adjust the classifier to pass through the low-density region between clusters. The synergistic learning between above regularizations during the training process enhances EEG decoding performance across subjects.

RESULTS

The effectiveness of the ADFR framework was evaluated on two public motor imagery (MI)-based EEG datasets: BCI Competition III dataset 4a and BCI Competition IV dataset 2a. In terms of average accuracy, ADFR achieved improvements of 3.0% and 2.1%, respectively, over the state-of-the-art methods on these datasets.

CONCLUSIONS

The promising results highlight the effectiveness of the ADFR algorithm for EEG decoding and show its potential for practical applications.

A Lightweight Network with Domain Adaptation for Motor Imagery Recognition

Brain-computer interfaces (BCI) are an effective tool for recognizing motor imagery and have been widely applied in the motor control and assistive operation domains. However, traditional intention-recognition methods face several challenges, such as prolonged training times and limited cross-subject adaptability, which restrict their practical application. This paper proposes an innovative method that combines a lightweight convolutional neural network (CNN) with domain adaptation. A lightweight feature extraction module is designed to extract key features from both the source and target domains, effectively reducing the model's parameters and improving the real-time performance and computational efficiency. To address differences in sample distributions, a domain adaptation strategy is introduced to optimize the feature alignment. Furthermore, domain adversarial training is employed to promote the learning of domain-invariant features, significantly enhancing the model's cross-subject generalization ability. The proposed method was evaluated on an fNIRS motor imagery dataset, achieving an average accuracy of 87.76% in a three-class classification task. Additionally, lightweight experiments were conducted from two perspectives: model structure optimization and data feature selection. The results demonstrated the potential advantages of this method for practical applications in motor imagery recognition systems.

Improving subject transfer in EEG classification with divergence estimation

Objective. Classification models for electroencephalogram (EEG) data show a large decrease in performance when evaluated on unseen test subjects. We improve performance using new regularization techniques during model training.Approach. We propose several graphical models to describe an EEG classification task. From each model, we identify statistical relationships that should hold true in an idealized training scenario (with infinite data and a globally-optimal model) but that may not hold in practice. We design regularization penalties to enforce these relationships in two stages. First, we identify suitable proxy quantities (divergences such as Mutual Information and Wasserstein-1) that can be used to measure statistical independence and dependence relationships. Second, we provide algorithms to efficiently estimate these quantities during training using secondary neural network models.Main results. We conduct extensive computational experiments using a large benchmark EEG dataset, comparing our proposed techniques with a baseline method that uses an adversarial classifier. We first show the performance of each method across a wide range of hyperparameters, demonstrating that each method can be easily tuned to yield significant benefits over an unregularized model. We show that, using ideal hyperparameters for all methods, our first technique gives significantly better performance than the baseline regularization technique. We also show that, across hyperparameters, our second technique gives significantly more stable performance than the baseline. The proposed methods require only a small computational cost at training time that is equivalent to the cost of the baseline.Significance. The high variability in signal distribution between subjects means that typical approaches to EEG signal modeling often require time-intensive calibration for each user, and even re-calibration before every use. By improving the performance of population models in the most stringent case of zero-shot subject transfer, we may help reduce or eliminate the need for model calibration.

Site-Invariant Meta-Modulation Learning for Multisite Autism Spectrum Disorders Diagnosis

Large amounts of fMRI data are essential to building generalized predictive models for brain disease diagnosis. In order to conduct extensive data analysis, it is often necessary to gather data from multiple organizations. However, the site variation inherent in multisite resting-state functional magnetic resonance imaging (rs-fMRI) leads to unfavorable heterogeneity in data distribution, negatively impacting the identification of biomarkers and the diagnostic decision. Several existing methods have alleviated this shift of domain distribution (i.e., multisite problem). Statistical tuning schemes directly regress out site disparity factors from the data prior to model training. Such methods have a limitation in processing data each time through variance estimation according to the added site. In the model adjustment approaches, domain adaptation (DA) methods adjust the features or models of the source domain according to the target domain during model training. Thus, it is inevitable that it needs updating model parameters according to the samples of a target site, causing great limitations in practical applicability. Meanwhile, the approach of domain generalization (DG) aims to create a universal model that can be quickly adapted to multiple domains. In this study, we propose a novel framework for disease diagnosis that alleviates the multisite problem by adaptively calibrating site-specific features into site-invariant features. Specifically, it applies directly to samples from unseen sites without the need for fine-tuning. With a learning-to-learn strategy that learns how to calibrate the features under the various domain shift environments, our novel modulation mechanism extracts site-invariant features. In our experiments over the Autism Brain Imaging Data Exchange (ABIDE I and II) dataset, we validated the generalization ability of the proposed network by improving diagnostic accuracy in both seen and unseen multisite samples.

Attention Induced Dual Convolutional-Capsule Network (AIDC-CN): A deep learning framework for motor imagery classification

In recent times, Electroencephalography (EEG)-based motor imagery (MI) decoding has garnered significant attention due to its extensive applicability in healthcare, including areas such as assistive robotics and rehabilitation engineering. Nevertheless, the decoding of EEG signals presents considerable challenges owing to their inherent complexity, non-stationary characteristics, and low signal-to-noise ratio. Notably, deep learning-based classifiers have emerged as a prominent focus for addressing the EEG signal decoding process. This study introduces a novel deep learning classifier named the Attention Induced Dual Convolutional-Capsule Network (AIDC-CN) with the specific aim of accurately categorizing various motor imagination class labels. To enhance the classifier's performance, a dual feature extraction approach leveraging spectrogram and brain connectivity networks has been employed, diversifying the feature set in the classification task. The main highlights of the proposed AIDC-CN classifier includes the introduction of a dual convolution layer to handle the brain connectivity and spectrogram features, addition of a novel self-attention module (SAM) to accentuate the relevant parts of the convolved spectrogram features, introduction of a new cross-attention module (CAM) to refine the outputs obtained from the dual convolution layers and incorporation of a Gaussian Error Linear Unit (GELU) based dynamic routing algorithm to strengthen the coupling among the primary and secondary capsule layers. Performance analysis undertaken on four public data sets depict the superior performance of the proposed model with respect to the state-of-the-art techniques. The code for this model is available at https://github.com/RiteshSurChowdhury/AIDC-CN.

MACNet: A Multidimensional Attention-Based Convolutional Neural Network for Lower-Limb Motor Imagery Classification

Decoding lower-limb motor imagery (MI) is highly important in brain-computer interfaces (BCIs) and rehabilitation engineering. However, it is challenging to classify lower-limb MI from electroencephalogram (EEG) signals, because lower-limb motions (LLMs) including MI are excessively close to physiological representations in the human brain and generate low-quality EEG signals. To address this challenge, this paper proposes a multidimensional attention-based convolutional neural network (CNN), termed MACNet, which is specifically designed for lower-limb MI classification. MACNet integrates a temporal refining module and an attention-enhanced convolutional module by leveraging the local and global feature representation abilities of CNNs and attention mechanisms. The temporal refining module adaptively investigates critical information from each electrode channel to refine EEG signals along the temporal dimension. The attention-enhanced convolutional module extracts temporal and spatial features while refining the feature maps across the channel and spatial dimensions. Owing to the scarcity of public datasets available for lower-limb MI, a specified lower-limb MI dataset involving four routine LLMs is built, consisting of 10 subjects over 20 sessions. Comparison experiments and ablation studies are conducted on this dataset and a public BCI Competition IV 2a EEG dataset. The experimental results show that MACNet achieves state-of-the-art performance and outperforms alternative models for the subject-specific mode. Visualization analysis reveals the excellent feature learning capabilities of MACNet and the potential relationship between lower-limb MI and brain activity. The effectiveness and generalizability of MACNet are verified.

A class alignment network based on self-attention for cross-subject EEG classification

Due to the inherent variability in EEG signals across different individuals, domain adaptation and adversarial learning strategies are being progressively utilized to develop subject-specific classification models by leveraging data from other subjects. These approaches primarily focus on domain alignment and tend to overlook the critical task-specific class boundaries. This oversight can result in weak correlation between the extracted features and categories. To address these challenges, we propose a novel model that uses the known information from multiple subjects to bolster EEG classification for an individual subject through adversarial learning strategies. Our method begins by extracting both shallow and attention-driven deep features from EEG signals. Subsequently, we employ a class discriminator to encourage the same-class features from different domains to converge while ensuring that the different-class features diverge. This is achieved using our proposed discrimination loss function, which is designed to minimize the feature distance for samples of the same class across different domains while maximizing it for those from different classes. Additionally, our model incorporates two parallel classifiers that are harmonious yet distinct and jointly contribute to decision-making. Extensive testing on two publicly available EEG datasets has validated our model's efficacy and superiority.

A Cross-Attention-Based Class Alignment Network for Cross-Subject EEG Classification in a Heterogeneous Space

BACKGROUND

Domain adaptation (DA) techniques have emerged as a pivotal strategy in addressing the challenges of cross-subject classification. However, traditional DA methods are inherently limited by the assumption of a homogeneous space, requiring that the source and target domains share identical feature dimensions and label sets, which is often impractical in real-world applications. Therefore, effectively addressing the challenge of EEG classification under heterogeneous spaces has emerged as a crucial research topic.

METHODS

We present a comprehensive framework that addresses the challenges of heterogeneous spaces by implementing a cross-domain class alignment strategy. We innovatively construct a cross-encoder to effectively capture the intricate dependencies between data across domains. We also introduce a tailored class discriminator accompanied by a corresponding loss function. By optimizing the loss function, we facilitate the aggregation of features with corresponding classes between the source and target domains, while ensuring that features from non-corresponding classes are dispersed.

RESULTS

Extensive experiments were conducted on two publicly available EEG datasets. Compared to advanced methods that combine label alignment with transfer learning, our method demonstrated superior performance across five heterogeneous space scenarios. Notably, in four heterogeneous label space scenarios, our method outperformed the advanced methods by an average of 7.8%. Moreover, in complex scenarios involving both heterogeneous label spaces and heterogeneous feature spaces, our method outperformed the state-of-the-art methods by an average of 4.1%.

CONCLUSIONS

This paper presents an efficient model for cross-subject EEG classification under heterogeneous spaces, which significantly addresses the challenges of EEG classification within heterogeneous spaces, thereby opening up new perspectives and avenues for research in related fields.

Unsupervised and semi-supervised domain adaptation networks considering both global knowledge and prototype-based local class information for Motor Imagery Classification

The non-stationarity of EEG signals results in variability across sessions, impeding model building and data sharing. In this paper, we propose a domain adaptation method called GPL, which simultaneously considers global knowledge and prototype-based local class information to enhance the classification accuracy of motor imagery signals. Depending on the amount of labeled data available in the target domain, the method is implemented in both unsupervised and semi-supervised versions. Specifically, at the global level, we employ the maximum mean difference (MMD) loss to globally constrain the feature space, achieving comprehensive alignment. In the context of class-level operations, we propose two memory banks designed to accommodate class prototypes in each domain and constrain feature embeddings by applying two prototype-based contrastive losses. The source contrastive loss is used to organize source features spatially based on categories, thereby reconciling inter-class and intra-class relationships, while the interactive contrastive loss is employed to facilitate cross-domain information interaction. Simultaneously, in unsupervised scenarios, to mitigate the adverse effects of excessive pseudo-labels, we introduce an entropy-aware strategy that dynamically evaluates the confidence level of target data and personalized constraints on the participation of interactive contrastive loss. To validate our approach, extensive experiments were conducted on a highly regarded public EEG dataset, namely Dataset IIa of the BCI Competition IV, as well as a large-scale EEG dataset called GigaDB. The experiments yielded average classification accuracies of 86.03% and 84.22% respectively. These results demonstrate that our method is an effective EEG decoding model, conducive to advancing the development of motor imagery brain-computer interfaces. The architecture proposed in this study and the code for data partitioning can be found at https://github.com/zhangdx21/GPL.

Session-independent subject-adaptive mental imagery BCI using selective filter-bank adaptive Riemannian features

The brain-computer interfaces (BCIs) facilitate the users to exploit information encoded in neural signals, specifically electroencephalogram (EEG), to control devices and for neural rehabilitation. Mental imagery (MI)-driven BCI predicts the user's pre-meditated mental objectives, which could be deployed as command signals. This paper presents a novel learning-based framework for classifying MI tasks using EEG-based BCI. In particular, our work focuses on the variation in inter-session data and the extraction of multi-spectral user-tailored features for robust performance. Thus, the goal is to create a calibration-free subject-adaptive learning framework for various mental imagery tasks not restricted to motor imagery alone. In this regard, critical spectral bands and the best temporal window are first selected from the EEG training trials of the subject based on the Riemannian user learning distance metric (Dscore) that checks for distinct and stable patterns. The filtered covariance matrices of the EEG trials in each spectral band are then transformed towards a reference covariance matrix using the Riemannian transfer learning, enabling the different sessions to be comparable. The evaluation of our proposed Selective Time-window and Multi-scale Filter-Bank with Adaptive Riemannian (STFB-AR) features on four public datasets, including disabled subjects, showed around 15% and 8% improvement in mean accuracy over baseline and fixed filter-bank models, respectively.

Adversarial Spatiotemporal Contrastive Learning for Electrocardiogram Signals

Extracting invariant representations in unlabeled electrocardiogram (ECG) signals is a challenge for deep neural networks (DNNs). Contrastive learning is a promising method for unsupervised learning. However, it should improve its robustness to noise and learn the spatiotemporal and semantic representations of categories, just like cardiologists. This article proposes a patient-level adversarial spatiotemporal contrastive learning (ASTCL) framework, which includes ECG augmentations, an adversarial module, and a spatiotemporal contrastive module. Based on the ECG noise attributes, two distinct but effective ECG augmentations, ECG noise enhancement, and ECG noise denoising, are introduced. These methods are beneficial for ASTCL to enhance the robustness of the DNN to noise. This article proposes a self-supervised task to increase the antiperturbation ability. This task is represented as a game between the discriminator and encoder in the adversarial module, which pulls the extracted representations into the shared distribution between the positive pairs to discard the perturbation representations and learn the invariant representations. The spatiotemporal contrastive module combines spatiotemporal prediction and patient discrimination to learn the spatiotemporal and semantic representations of categories. To learn category representations effectively, this article only uses patient-level positive pairs and alternately uses the predictor and the stop-gradient to avoid model collapse. To verify the effectiveness of the proposed method, various groups of experiments are conducted on four ECG benchmark datasets and one clinical dataset compared with the state-of-the-art methods. Experimental results showed that the proposed method outperforms the state-of-the-art methods.

Cross-Channel Specific-Mutual Feature Transfer Learning for Motor Imagery EEG Signals Decoding

In recent years, with the rapid development of deep learning, various deep learning frameworks have been widely used in brain-computer interface (BCI) research for decoding motor imagery (MI) electroencephalogram (EEG) signals to understand brain activity accurately. The electrodes, however, record the mixed activities of neurons. If different features are directly embedded in the same feature space, the specific and mutual features of different neuron regions are not considered, which will reduce the expression ability of the feature itself. We propose a cross-channel specific-mutual feature transfer learning (CCSM-FT) network model to solve this problem. The multibranch network extracts the specific and mutual features of brain's multiregion signals. Effective training tricks are used to maximize the distinction between the two kinds of features. Suitable training tricks can also improve the effectiveness of the algorithm compared with novel models. Finally, we transfer two kinds of features to explore the potential of mutual and specific features to enhance the expressive power of the feature and use the auxiliary set to improve identification performance. The experimental results show that the network has a better classification effect in the BCI Competition IV-2a and the HGD datasets.

Continual learning for seizure prediction via memory projection strategy

Despite extensive algorithms for epilepsy prediction via machine learning, most models are tailored for offline scenarios and cannot handle actual scenarios where data changes over time. Catastrophic forgetting(CF) for learned electroencephalogram(EEG) data occurs when EEG changes dynamically in the clinical setting. This paper implements a continual learning(CL) strategy Memory Projection(MP) for epilepsy prediction, which can be combined with other algorithms to avoid CF. Such a strategy enables the model to learn EEG data from each patient in dynamic subspaces with weak correlation layer by layer to minimize interference and promote knowledge transfer. Regularization Loss Reconstruction Algorithm and Matrix Dimensionality Reduction Algorithm are introduced into the core of MP. Experimental results show that MP exhibits excellent performance and low forgetting rates in sequential learning of seizure prediction. The forgetting rate of accuracy and sensitivity under multiple experiments are below 5%. When learning from multi-center datasets, the forgetting rates for accuracy and sensitivity decrease to 0.65% and 1.86%, making it comparable to state-of-the-art CL strategies. Through ablation experiments, we have analyzed that MP can operate with minimal storage and computational cost, which demonstrates practical potential for seizure prediction in clinical scenarios.

Multiclass motor imagery classification with Riemannian geometry and temporal-spectral selection

Motor imagery (MI) based brain-computer interfaces (BCIs) decode the users' intentions from electroencephalography (EEG) to achieve information control and interaction between the brain and external devices. In this paper, firstly, we apply Riemannian geometry to the covariance matrix extracted by spatial filtering to obtain robust and distinct features. Then, a multiscale temporal-spectral segmentation scheme is developed to enrich the feature dimensionality. In order to determine the optimal feature configurations, we utilize a linear learning-based temporal window and spectral band (TWSB) selection method to evaluate the feature contributions, which efficiently reduces the redundant features and improves the decoding efficiency without excessive loss of accuracy. Finally, support vector machines are used to predict the classification labels based on the selected MI features. To evaluate the performance of our model, we test it on the publicly available BCI Competition IV dataset 2a and 2b. The results show that the method has an average accuracy of 79.1% and 83.1%, which outperforms other existing methods. Using TWSB feature selection instead of selecting all features improves the accuracy by up to about 6%. Moreover, the TWSB selection method can effectively reduce the computational burden. We believe that the framework reveals more interpretable feature information of motor imagery EEG signals, provides neural responses discriminative with high accuracy, and facilitates the performance of real-time MI-BCI.

User Evaluation of a Shared Robot Control System Combining BCI and Eye Tracking in a Portable Augmented Reality User Interface

This study evaluates an innovative control approach to assistive robotics by integrating brain-computer interface (BCI) technology and eye tracking into a shared control system for a mobile augmented reality user interface. Aimed at enhancing the autonomy of individuals with physical disabilities, particularly those with impaired motor function due to conditions such as stroke, the system utilizes BCI to interpret user intentions from electroencephalography signals and eye tracking to identify the object of focus, thus refining control commands. This integration seeks to create a more intuitive and responsive assistive robot control strategy. The real-world usability was evaluated, demonstrating significant potential to improve autonomy for individuals with severe motor impairments. The control system was compared with an eye-tracking-based alternative to identify areas needing improvement. Although BCI achieved an acceptable success rate of 0.83 in the final phase, eye tracking was more effective with a perfect success rate and consistently lower completion times (p<0.001). The user experience responses favored eye tracking in 11 out of 26 questions, with no significant differences in the remaining questions, and subjective fatigue was higher with BCI use (p=0.04). While BCI performance lagged behind eye tracking, the user evaluation supports the validity of our control strategy, showing that it could be deployed in real-world conditions and suggesting a pathway for further advancements.

EEG-Based Motor Imagery Recognition Framework via Multisubject Dynamic Transfer and Iterative Self-Training

A robust decoding model that can efficiently deal with the subject and period variation is urgently needed to apply the brain-computer interface (BCI) system. The performance of most electroencephalogram (EEG) decoding models depends on the characteristics of specific subjects and periods, which require calibration and training with annotated data prior to application. However, this situation will become unacceptable as it would be difficult for subjects to collect data for an extended period, especially in the rehabilitation process of disability based on motor imagery (MI). To address this issue, we propose an unsupervised domain adaptation framework called iterative self-training multisubject domain adaptation (ISMDA) that focuses on the offline MI task. First, the feature extractor is purposefully designed to map the EEG to a latent space of discriminative representations. Second, the attention module based on dynamic transfer matches the source domain and target domain samples with a higher coincidence degree in latent space. Then, an independent classifier oriented to the target domain is employed in the first stage of the iterative training process to cluster the samples of the target domain through similarity. Finally, a pseudolabel algorithm based on certainty and confidence is employed in the second stage of the iterative training process to adequately calibrate the error between prediction and empirical probabilities. To evaluate the effectiveness of the model, extensive testing has been performed on three publicly available MI datasets, the BCI IV IIa, the High gamma dataset, and Kwon et al. datasets. The proposed method achieved 69.51%, 82.38%, and 90.98% cross-subject classification accuracy on the three datasets, which outperforms the current state-of-the-art offline algorithms. Meanwhile, all results demonstrated that the proposed method could address the main challenges of the offline MI paradigm.

Unsupervised heterogeneous domain adaptation for EEG classification

Objective.Domain adaptation has been recognized as a potent solution to the challenge of limited training data for electroencephalography (EEG) classification tasks. Existing studies primarily focus on homogeneous environments, however, the heterogeneous properties of EEG data arising from device diversity cannot be overlooked. This motivates the development of heterogeneous domain adaptation methods that can fully exploit the knowledge from an auxiliary heterogeneous domain for EEG classification.Approach.In this article, we propose a novel model named informative representation fusion (IRF) to tackle the problem of unsupervised heterogeneous domain adaptation in the context of EEG data. In IRF, we consider different perspectives of data, i.e. independent identically distributed (iid) and non-iid, to learn different representations. Specifically, from the non-iid perspective, IRF models high-order correlations among data by hypergraphs and develops hypergraph encoders to obtain data representations of each domain. From the non-iid perspective, by applying multi-layer perceptron networks to the source and target domain data, we achieve another type of representation for both domains. Subsequently, an attention mechanism is used to fuse these two types of representations to yield informative features. To learn transferable representations, the maximum mean discrepancy is utilized to align the distributions of the source and target domains based on the fused features.Main results.Experimental results on several real-world datasets demonstrate the effectiveness of the proposed model.Significance.This article handles an EEG classification situation where the source and target EEG data lie in different spaces, and what's more, under an unsupervised learning setting. This situation is practical in the real world but barely studied in the literature. The proposed model achieves high classification accuracy, and this study is important for the commercial applications of EEG-based BCIs.

Harnessing Few-Shot Learning for EEG signal classification: a survey of state-of-the-art techniques and future directions

This paper presents a systematic literature review, providing a comprehensive taxonomy of Data Augmentation (DA), Transfer Learning (TL), and Self-Supervised Learning (SSL) techniques within the context of Few-Shot Learning (FSL) for EEG signal classification. EEG signals have shown significant potential in various paradigms, including Motor Imagery, Emotion Recognition, Visual Evoked Potentials, Steady-State Visually Evoked Potentials, Rapid Serial Visual Presentation, Event-Related Potentials, and Mental Workload. However, challenges such as limited labeled data, noise, and inter/intra-subject variability have impeded the effectiveness of traditional machine learning (ML) and deep learning (DL) models. This review methodically explores how FSL approaches, incorporating DA, TL, and SSL, can address these challenges and enhance classification performance in specific EEG paradigms. It also delves into the open research challenges related to these techniques in EEG signal classification. Specifically, the review examines the identification of DA strategies tailored to various EEG paradigms, the creation of TL architectures for efficient knowledge transfer, and the formulation of SSL methods for unsupervised representation learning from EEG data. Addressing these challenges is crucial for enhancing the efficacy and robustness of FSL-based EEG signal classification. By presenting a structured taxonomy of FSL techniques and discussing the associated research challenges, this systematic review offers valuable insights for future investigations in EEG signal classification. The findings aim to guide and inspire researchers, promoting advancements in applying FSL methodologies for improved EEG signal analysis and classification in real-world settings.

Cross-subject EEG-based Motor Imagery Recognition for Patient's Rehabilitation

Motor imagery (MI), a kind of psychological representation without actual action, has garnered increasing attention in rehabilitation. However, the inherent differences between patients and healthy persons hinder rehabilitation by reducing the accuracy of cross-subject MI recognition. Although unsupervised domain adaptation (UDA) methods have mitigated individual differences, they still suffer from challenges in terms of selecting confusing source domains and accurately classifying MI samples at the boundary. To address these challenges, we propose a novel UDA framework with a causal graphical model and label similarity clustering. The causal graphical model is employed to estimate the similarity of EEG signals, enabling the causal selection to effectively avoid confusing healthy persons' data. In addition, label similarity clustering mechanism is utilized to establish a distinct boundary, thereby enhancing the classification accuracy. The experimental results demonstrate that our approach outperforms baseline 10.10% and 16.27% on BCI IV-2a&2b, separately. MI is expected to aid rehabilitation through precise recognition and active support.

Proxy-based Masking Module for Revealing Relevance of Characteristics in Motor Imagery

Brain-computer interface (BCI) has been developed for communication between users and external devices by reflecting users' status and intentions. Motor imagery (MI) is one of the BCI paradigms for controlling external devices by imagining muscle movements. MI-based EEG signals generally tend to contain signals with sparse MI characteristics (sparse MI signals). When conducting domain adaptation (DA) on MI signals with sparse MI signals, it could interrupt the training process. In this paper, we proposed the proxy-based masking module (PMM) for masking sparse MI signals within MI signals. The proposed module was designed to suppress the amplitude of sparse MI signals using the negative similarity-based mask generated between the proxy of rest signals and the feature vectors of MI signals. We attached our proposed module to the conventional DA methods (i.e., the DJDAN, the MAAN, and the DRDA) to verify the effectiveness in the cross-subject environment on dataset 2a of BCI competition IV. When our proposed module was attached to each conventional DA method, the average accuracy was improved by much as 4.67 %, 0.76 %, and 1.72 %, respectively. Hence, we demonstrated that our proposed module could emphasize the information related to MI characteristics. The code of our implementation is accessible on GitHub.1.

Semi-supervised multi-source transfer learning for cross-subject EEG motor imagery classification

Electroencephalogram (EEG) motor imagery (MI) classification refers to the use of EEG signals to identify and classify subjects' motor imagery activities; this task has received increasing attention with the development of brain-computer interfaces (BCIs). However, the collection of EEG data is usually time-consuming and labor-intensive, which makes it difficult to obtain sufficient labeled data from the new subject to train a new model. Moreover, the EEG signals of different individuals exhibit significant differences, leading to a significant drop in the performance of a model trained on the existing subjects when directly classifying EEG signals acquired from new subjects. Therefore, it is crucial to make full use of the EEG data of the existing subjects and the unlabeled EEG data of the new target subject to improve the MI classification performance achieved for the target subject. This research study proposes a semi-supervised multi-source transfer (SSMT) learning model to address the above problems; the model learns informative and domain-invariant representations to address cross-subject MI-EEG classification tasks. In particular, a dynamic transferred weighting schema is presented to obtain the final predictions by integrating the weighted features derived from multi-source domains. The average accuracies achieved on two publicly available EEG datasets reach 83.57 % and 85.09 % , respectively, validating the effectiveness of the SSMT process. The SSMT process reveals the importance of informative and domain-invariant representations in MI classification tasks, as they make full use of the domain-invariant information acquired from each subject.

Bidirectional feature pyramid attention-based temporal convolutional network model for motor imagery electroencephalogram classification

Introduction

As an interactive method gaining popularity, brain-computer interfaces (BCIs) aim to facilitate communication between the brain and external devices. Among the various research topics in BCIs, the classification of motor imagery using electroencephalography (EEG) signals has the potential to greatly improve the quality of life for people with disabilities.

Methods

This technology assists them in controlling computers or other devices like prosthetic limbs, wheelchairs, and drones. However, the current performance of EEG signal decoding is not sufficient for real-world applications based on Motor Imagery EEG (MI-EEG). To address this issue, this study proposes an attention-based bidirectional feature pyramid temporal convolutional network model for the classification task of MI-EEG. The model incorporates a multi-head self-attention mechanism to weigh significant features in the MI-EEG signals. It also utilizes a temporal convolution network (TCN) to separate high-level temporal features. The signals are enhanced using the sliding-window technique, and channel and time-domain information of the MI-EEG signals is extracted through convolution.

Results

Additionally, a bidirectional feature pyramid structure is employed to implement attention mechanisms across different scales and multiple frequency bands of the MI-EEG signals. The performance of our model is evaluated on the BCI Competition IV-2a dataset and the BCI Competition IV-2b dataset, and the results showed that our model outperformed the state-of-the-art baseline model, with an accuracy of 87.5 and 86.3% for the subject-dependent, respectively.

Discussion

In conclusion, the BFATCNet model offers a novel approach for EEG-based motor imagery classification in BCIs, effectively capturing relevant features through attention mechanisms and temporal convolutional networks. Its superior performance on the BCI Competition IV-2a and IV-2b datasets highlights its potential for real-world applications. However, its performance on other datasets may vary, necessitating further research on data augmentation techniques and integration with multiple modalities to enhance interpretability and generalization. Additionally, reducing computational complexity for real-time applications is an important area for future work.

An in-depth survey on Deep Learning-based Motor Imagery Electroencephalogram (EEG) classification

Electroencephalogram (EEG)-based Brain-Computer Interfaces (BCIs) build a communication path between human brain and external devices. Among EEG-based BCI paradigms, the most commonly used one is motor imagery (MI). As a hot research topic, MI EEG-based BCI has largely contributed to medical fields and smart home industry. However, because of the low signal-to-noise ratio (SNR) and the non-stationary characteristic of EEG data, it is difficult to correctly classify different types of MI-EEG signals. Recently, the advances in Deep Learning (DL) significantly facilitate the development of MI EEG-based BCIs. In this paper, we provide a systematic survey of DL-based MI-EEG classification methods. Specifically, we first comprehensively discuss several important aspects of DL-based MI-EEG classification, covering input formulations, network architectures, public datasets, etc. Then, we summarize problems in model performance comparison and give guidelines to future studies for fair performance comparison. Next, we fairly evaluate the representative DL-based models using source code released by the authors and meticulously analyse the evaluation results. By performing ablation study on the network architecture, we found that (1) effective feature fusion is indispensable for multi-stream CNN-based models. (2) LSTM should be combined with spatial feature extraction techniques to obtain good classification performance. (3) the use of dropout contributes little to improving the model performance, and that (4) adding fully connected layers to the models significantly increases their parameters but it might not improve their performance. Finally, we raise several open issues in MI-EEG classification and provide possible future research directions.

EEG decoding for datasets with heterogenous electrode configurations using transfer learning graph neural networks

Objective. Brain-machine interfacing (BMI) has greatly benefited from adopting machine learning methods for feature learning that require extensive data for training, which are often unavailable from a single dataset. Yet, it is difficult to combine data across labs or even data within the same lab collected over the years due to the variation in recording equipment and electrode layouts resulting in shifts in data distribution, changes in data dimensionality, and altered identity of data dimensions. Our objective is to overcome this limitation and learn from many different and diverse datasets across labs with different experimental protocols.Approach. To tackle the domain adaptation problem, we developed a novel machine learning framework combining graph neural networks (GNNs) and transfer learning methodologies for non-invasive motor imagery (MI) EEG decoding, as an example of BMI. Empirically, we focus on the challenges of learning from EEG data with different electrode layouts and varying numbers of electrodes. We utilize three MI EEG databases collected using very different numbers of EEG sensors (from 22 channels to 64) and layouts (from custom layouts to 10-20).Main results. Our model achieved the highest accuracy with lower standard deviations on the testing datasets. This indicates that the GNN-based transfer learning framework can effectively aggregate knowledge from multiple datasets with different electrode layouts, leading to improved generalization in subject-independent MI EEG classification.Significance. The findings of this study have important implications for brain-computer-interface research, as they highlight a promising method for overcoming the limitations posed by non-unified experimental setups. By enabling the integration of diverse datasets with varying electrode layouts, our proposed approach can help advance the development and application of BMI technologies.

Cross-dataset transfer learning for motor imagery signal classification via multi-task learning and pre-training

Objective.Deep learning (DL) models have been proven to be effective in decoding motor imagery (MI) signals in Electroencephalogram (EEG) data. However, DL models' success relies heavily on large amounts of training data, whereas EEG data collection is laborious and time-consuming. Recently, cross-dataset transfer learning has emerged as a promising approach to meet the data requirements of DL models. Nevertheless, transferring knowledge across datasets involving different MI tasks remains a significant challenge in cross-dataset transfer learning, limiting the full utilization of valuable data resources.

APPROACH

This study proposes a pre-training-based cross-dataset transfer learning method inspired by Hard Parameter Sharing in multi-task learning. Different datasets with distinct MI paradigms are considered as different tasks, classified with shared feature extraction layers and individual task-specific layers to allow cross-dataset classification with one unified model. Then, Pre-training and fine-tuning are employed to transfer knowledge across datasets. We also designed four fine-tuning schemes and conducted extensive experiments on them.

MAIN RESULTS

The results showed that compared to models without pre-training, models with pre-training achieved a maximum increase in accuracy of 7.76%. Moreover, when limited training data were available, the pre-training method significantly improved DL model's accuracy by 27.34% at most. The experiments also revealed that pre-trained models exhibit faster convergence and remarkable robustness. The training time per subject could be reduced by up to 102.83 s, and the variance of classification accuracy decreased by 75.22% at best.

SIGNIFICANCE

This study represents the first comprehensive investigation of the cross-dataset transfer learning method between two datasets with different MI tasks. The proposed pre-training method requires only minimal fine-tuning data when applying DL models to new MI paradigms, making MI-Brain-computer interface more practical and user-friendly.

MI-DAGSC: A domain adaptation approach incorporating comprehensive information from MI-EEG signals

Non-stationarity of EEG signals leads to high variability between subjects, making it challenging to directly use data from other subjects (source domain) for the classifier in the current subject (target domain). In this study, we propose MI-DAGSC to address domain adaptation challenges in EEG-based motor imagery (MI) decoding. By combining domain-level information, class-level information, and inter-sample structure information, our model effectively aligns the feature distributions of source and target domains. This work is an extension of our previous domain adaptation work MI-DABAN (Li et al., 2023). Based on MI-DABAN, MI-DAGSC designs Sample-Feature Blocks (SFBs) and Graph Convolution Blocks (GCBs) to focus on intra-sample and inter-sample information. The synergistic integration of SFBs and GCBs enable the model to capture comprehensive information and understand the relationship between samples, thus improving representation learning. Furthermore, we introduce a triplet loss to enhance the alignment and compactness of feature representations. Extensive experiments on real EEG datasets demonstrate the effectiveness of MI-DAGSC, confirming that our method makes a valuable contribution to the MI-EEG decoding. Moreover, it holds great potential for various applications in brain-computer interface systems and neuroscience research. And the code of the proposed architecture in this study is available under https://github.com/zhangdx21/MI-DAGSC.

EEG Signal Epilepsy Detection With a Weighted Neighbor Graph Representation and Two-Stream Graph-Based Framework

Epilepsy is one of the most common neurological diseases. Clinically, epileptic seizure detection is usually performed by analyzing electroencephalography (EEG) signals. At present, deep learning models have been widely used for single-channel EEG signal epilepsy detection, but this method is difficult to explain the classification results. Researchers have attempted to solve interpretive problems by combining graph representation of EEG signals with graph neural network models. Recently, the combination of graph representations and graph neural network (GNN) models has been increasingly applied to single-channel epilepsy detection. By this methodology, the raw EEG signal is transformed to its graph representation, and a GNN model is used to learn latent features and classify whether the data indicates an epileptic seizure episode. However, existing methods are faced with two major challenges. First, existing graph representations tend to have high time complexity as they generally require each vertex to traverse all other vertices to construct a graph structure. Some of them also have high space complexity for being dense. Second, while separate graph representations can be derived from a single-channel EEG signal in both time and frequency domains, existing GNN models for epilepsy detection can learn from a single graph representation, which makes it hard to let the information from the two domains complement each other. For addressing these challenges, we propose a Weighted Neighbour Graph (WNG) representation for EEG signals. Reducing the redundant edges of the existing graph, WNG can be both time and space-efficient, and as informative as its less efficient counterparts. We then propose a two-stream graph-based framework to simultaneously learn features from WNG in both time and frequency domain. Extensive experiments demonstrate the effectiveness and efficiency of the proposed methods.

LMDA-Net:A lightweight multi-dimensional attention network for general EEG-based brain-computer interfaces and interpretability

Electroencephalography (EEG)-based brain-computer interfaces (BCIs) pose a challenge for decoding due to their low spatial resolution and signal-to-noise ratio. Typically, EEG-based recognition of activities and states involves the use of prior neuroscience knowledge to generate quantitative EEG features, which may limit BCI performance. Although neural network-based methods can effectively extract features, they often encounter issues such as poor generalization across datasets, high predicting volatility, and low model interpretability. To address these limitations, we propose a novel lightweight multi-dimensional attention network, called LMDA-Net. By incorporating two novel attention modules designed specifically for EEG signals, the channel attention module and the depth attention module, LMDA-Net is able to effectively integrate features from multiple dimensions, resulting in improved classification performance across various BCI tasks. LMDA-Net was evaluated on four high-impact public datasets, including motor imagery (MI) and P300-Speller, and was compared with other representative models. The experimental results demonstrate that LMDA-Net outperforms other representative methods in terms of classification accuracy and predicting volatility, achieving the highest accuracy in all datasets within 300 training epochs. Ablation experiments further confirm the effectiveness of the channel attention module and the depth attention module. To facilitate an in-depth understanding of the features extracted by LMDA-Net, we propose class-specific neural network feature interpretability algorithms that are suitable for evoked responses and endogenous activities. By mapping the output of the specific layer of LMDA-Net to the time or spatial domain through class activation maps, the resulting feature visualizations can provide interpretable analysis and establish connections with EEG time-spatial analysis in neuroscience. In summary, LMDA-Net shows great potential as a general decoding model for various EEG tasks.

MI-CAT: A transformer-based domain adaptation network for motor imagery classification

Due to its convenience and safety, electroencephalography (EEG) data is one of the most widely used signals in motor imagery (MI) brain-computer interfaces (BCIs). In recent years, methods based on deep learning have been widely applied to the field of BCIs, and some studies have gradually tried to apply Transformer to EEG signal decoding due to its superior global information focusing ability. However, EEG signals vary from subject to subject. Based on Transformer, how to effectively use data from other subjects (source domain) to improve the classification performance of a single subject (target domain) remains a challenge. To fill this gap, we propose a novel architecture called MI-CAT. The architecture innovatively utilizes Transformer's self-attention and cross-attention mechanisms to interact features to resolve differential distribution between different domains. Specifically, we adopt a patch embedding layer for the extracted source and target features to divide the features into multiple patches. Then, we comprehensively focus on the intra-domain and inter-domain features by stacked multiple Cross-Transformer Blocks (CTBs), which can adaptively conduct bidirectional knowledge transfer and information exchange between domains. Furthermore, we also utilize two non-shared domain-based attention blocks to efficiently capture domain-dependent information, optimizing the features extracted from the source and target domains to assist in feature alignment. To evaluate our method, we conduct extensive experiments on two real public EEG datasets, Dataset IIb and Dataset IIa, achieving competitive performance with an average classification accuracy of 85.26% and 76.81%, respectively. Experimental results demonstrate that our method is a powerful model for decoding EEG signals and facilitates the development of the Transformer for brain-computer interfaces (BCIs).

An extended clinical EEG dataset with 15,300 automatically labelled recordings for pathology decoding

Automated clinical EEG analysis using machine learning (ML) methods is a growing EEG research area. Previous studies on binary EEG pathology decoding have mainly used the Temple University Hospital (TUH) Abnormal EEG Corpus (TUAB) which contains approximately 3,000 manually labelled EEG recordings. To evaluate and eventually even improve the generalisation performance of machine learning methods for EEG pathology, decoding larger, publicly available datasets is required. A number of studies addressed the automatic labelling of large open-source datasets as an approach to create new datasets for EEG pathology decoding, but little is known about the extent to which training on larger, automatically labelled dataset affects decoding performances of established deep neural networks. In this study, we automatically created additional pathology labels for the Temple University Hospital (TUH) EEG Corpus (TUEG) based on the medical reports using a rule-based text classifier. We generated a dataset of 15,300 newly labelled recordings, which we call the TUH Abnormal Expansion EEG Corpus (TUABEX), and which is five times larger than the TUAB. Since the TUABEX contains more pathological (75%) than non-pathological (25%) recordings, we then selected a balanced subset of 8,879 recordings, the TUH Abnormal Expansion Balanced EEG Corpus (TUABEXB). To investigate how training on a larger, automatically labelled dataset affects the decoding performance of deep neural networks, we applied four established deep convolutional neural networks (ConvNets) to the task of pathological versus non-pathological classification and compared the performance of each architecture after training on different datasets. The results show that training on the automatically labelled TUABEXB dataset rather than training on the manually labelled TUAB dataset increases accuracies on TUABEXB and even for TUAB itself for some architectures. We argue that automatically labelling of large open-source datasets can be used to efficiently utilise the massive amount of EEG data stored in clinical archives. We make the proposed TUABEXB available open source and thus offer a new dataset for EEG machine learning research.

Effect of Lower Limb Exoskeleton on the Modulation of Neural Activity and Gait Classification

Neurorehabilitation with robotic devices requires a paradigm shift to enhance human-robot interaction. The coupling of robot assisted gait training (RAGT) with a brain-machine interface (BMI) represents an important step in this direction but requires better elucidation of the effect of RAGT on the user's neural modulation. Here, we investigated how different exoskeleton walking modes modify brain and muscular activity during exoskeleton assisted gait. We recorded electroencephalographic (EEG) and electromyographic (EMG) activity from ten healthy volunteers walking with an exoskeleton with three modes of user assistance (i.e., transparent, adaptive and full assistance) and during free overground gait. Results identified that exoskeleton walking (irrespective of the exoskeleton mode) induces a stronger modulation of central mid-line mu (8-13 Hz) and low-beta (14-20 Hz) rhythms compared to free overground walking. These modifications are accompanied by a significant re-organization of the EMG patterns in exoskeleton walking. On the other hand, we observed no significant differences in neural activity during exoskeleton walking with the different assistance levels. We subsequently implemented four gait classifiers based on deep neural networks trained on the EEG data during the different walking conditions. Our hypothesis was that exoskeleton modes could impact the creation of a BMI-driven RAGT. We demonstrated that all classifiers achieved an average accuracy of 84.13±3.49% in classifying swing and stance phases on their respective datasets. In addition, we demonstrated that the classifier trained on the transparent mode exoskeleton data can classify gait phases during adaptive and full modes with an accuracy of 78.3±4.8% , while the classifier trained on free overground walking data fails to classify the gait during exoskeleton walking (accuracy of 59.4±11.8% ). These findings provide important insights into the effect of robotic training on neural activity and contribute to the advancement of BMI technology for improving robotic gait rehabilitation therapy.

Feature separation and adversarial training for the patient-independent detection of epileptic seizures

An epileptic seizure is the external manifestation of abnormal neuronal discharges, which seriously affecting physical health. The pathogenesis of epilepsy is complex, and the types of epileptic seizures are diverse, resulting in significant variation in epileptic seizure data between subjects. If we feed epilepsy data from multiple patients directly into the model for training, it will lead to underfitting of the model. To overcome this problem, we propose a robust epileptic seizure detection model that effectively learns from multiple patients while eliminating the negative impact of the data distribution shift between patients. The model adopts a multi-level temporal-spectral feature extraction network to achieve feature extraction, a feature separation network to separate features into category-related and patient-related components, and an invariant feature extraction network to extract essential feature information related to categories. The proposed model is evaluated on the TUH dataset using leave-one-out cross-validation and achieves an average accuracy of 85.7%. The experimental results show that the proposed model is superior to the related literature and provides a valuable reference for the clinical application of epilepsy detection.

Bridging the BCI illiteracy gap: a subject-to-subject semantic style transfer for EEG-based motor imagery classification

Introduction

Brain-computer interfaces (BCIs) facilitate direct interaction between the human brain and computers, enabling individuals to control external devices through cognitive processes. Despite its potential, the problem of BCI illiteracy remains one of the major challenges due to inter-subject EEG variability, which hinders many users from effectively utilizing BCI systems. In this study, we propose a subject-to-subject semantic style transfer network (SSSTN) at the feature-level to address the BCI illiteracy problem in electroencephalogram (EEG)-based motor imagery (MI) classification tasks.

Methods

Our approach uses the continuous wavelet transform method to convert high-dimensional EEG data into images as input data. The SSSTN 1) trains a classifier for each subject, 2) transfers the distribution of class discrimination styles from the source subject (the best-performing subject for the classifier, i.e., BCI expert) to each subject of the target domain (the remaining subjects except the source subject, specifically BCI illiterates) through the proposed style loss, and applies a modified content loss to preserve the class-relevant semantic information of the target domain, and 3) finally merges the classifier predictions of both source and target subject using an ensemble technique.

Results and discussion

We evaluate the proposed method on the BCI Competition IV-2a and IV-2b datasets and demonstrate improved classification performance over existing methods, especially for BCI illiterate users. The ablation experiments and t-SNE visualizations further highlight the effectiveness of the proposed method in achieving meaningful feature-level semantic style transfer.

Cross-subject aesthetic preference recognition of Chinese dance posture using EEG

Due to the differences in knowledge, experience, background, and social influence, people have subjective characteristics in the process of dance aesthetic cognition. To explore the neural mechanism of the human brain in the process of dance aesthetic preference, and to find a more objective determining criterion for dance aesthetic preference, this paper constructs a cross-subject aesthetic preference recognition model of Chinese dance posture. Specifically, Dai nationality dance (a classic Chinese folk dance) was used to design dance posture materials, and an experimental paradigm for aesthetic preference of Chinese dance posture was built. Then, 91 subjects were recruited for the experiment, and their EEG signals were collected. Finally, the transfer learning method and convolutional neural networks were used to identify the aesthetic preference of the EEG signals. Experimental results have shown the feasibility of the proposed model, and the objective aesthetic measurement in dance appreciation has been implemented. Based on the classification model, the accuracy of aesthetic preference recognition is 79.74%. Moreover, the recognition accuracies of different brain regions, different hemispheres, and different model parameters were also verified by the ablation study. Additionally, the experimental results reflected the following two facts: (1) in the visual aesthetic processing of Chinese dance posture, the occipital and frontal lobes are more activated and participate in dance aesthetic preference; (2) the right brain is more involved in the visual aesthetic processing of Chinese dance posture, which is consistent with the common knowledge that the right brain is responsible for processing artistic activities.

Improved Domain Adaptation Network Based on Wasserstein Distance for Motor Imagery EEG Classification

Motor Imagery (MI) paradigm is critical in neural rehabilitation and gaming. Advances in brain-computer interface (BCI) technology have facilitated the detection of MI from electroencephalogram (EEG). Previous studies have proposed various EEG-based classification algorithms to identify the MI, however, the performance of prior models was limited due to the cross-subject heterogeneity in EEG data and the shortage of EEG data for training. Therefore, inspired by generative adversarial network (GAN), this study aims to propose an improved domain adaption network based on Wasserstein distance, which utilizes existing labeled data from multiple subjects (source domain) to improve the performance of MI classification on a single subject (target domain). Specifically, our proposed framework consists of three components, including a feature extractor, a domain discriminator, and a classifier. The feature extractor employs an attention mechanism and a variance layer to improve the discrimination of features extracted from different MI classes. Next, the domain discriminator adopts the Wasserstein matrix to measure the distance between source domain and target domain, and aligns the data distributions of source and target domain via adversarial learning strategy. Finally, the classifier uses the knowledge acquired from the source domain to predict the labels in the target domain. The proposed EEG-based MI classification framework was evaluated by two open-source datasets, the BCI Competition IV Datasets 2a and 2b. Our results demonstrated that the proposed framework could enhance the performance of EEG-based MI detection, achieving better classification results compared with several state-of-the-art algorithms. In conclusion, this study is promising in helping the neural rehabilitation of different neuropsychiatric diseases.

EEG Conformer: Convolutional Transformer for EEG Decoding and Visualization

Due to the limited perceptual field, convolutional neural networks (CNN) only extract local temporal features and may fail to capture long-term dependencies for EEG decoding. In this paper, we propose a compact Convolutional Transformer, named EEG Conformer, to encapsulate local and global features in a unified EEG classification framework. Specifically, the convolution module learns the low-level local features throughout the one-dimensional temporal and spatial convolution layers. The self-attention module is straightforwardly connected to extract the global correlation within the local temporal features. Subsequently, the simple classifier module based on fully-connected layers is followed to predict the categories for EEG signals. To enhance interpretability, we also devise a visualization strategy to project the class activation mapping onto the brain topography. Finally, we have conducted extensive experiments to evaluate our method on three public datasets in EEG-based motor imagery and emotion recognition paradigms. The experimental results show that our method achieves state-of-the-art performance and has great potential to be a new baseline for general EEG decoding. The code has been released in https://github.com/eeyhsong/EEG-Conformer.

Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer Interfaces

For practical motor imagery (MI) brain-computer interface (BCI) applications, generating a reliable model for a target subject with few MI trials is important since the data collection process is labour-intensive and expensive. In this paper, we address this issue by proposing a few-shot learning method called temporal episode relation learning (TERL). TERL models MI with only limited trials from the target subject by the ability to compare MI trials through episode-based training. It can be directly applied to a new user without being re-trained, which is vital to improve user experience and realize real-world MIBCI applications. We develop a new and effective approach where, unlike the original episode learning, the temporal pattern between trials in each episode is encoded during the learning to boost the classification performance. We also perform an online evaluation simulation, in addition to the offline analysis that the previous studies only conduct, to better understand the performance of different approaches in real-world scenario. Extensive experiments are completed on four publicly available MIBCI datasets to evaluate the proposed TERL. Results show that TERL outperforms baseline and recent state-of-the-art methods, demonstrating competitive performance for subject-specific MIBCI where few trials are available from a target subject and a considerable number of trials from other source subjects.

MI-DABAN: A dual-attention-based adversarial network for motor imagery classification

The brain-computer interface (BCI) based on motor imagery electroencephalography (EEG) is widely used because of its convenience and safety. However, due to the distributional disparity between EEG signals, data from other subjects cannot be used directly to train a subject-specific classifier. For efficient use of the labeled data, domain transfer learning and adversarial learning are gradually applied to BCI classification tasks. While these methods improve classification performance, they only align globally and ignore task-specific class boundaries, which may lead to the blurring of features near the classification boundaries. Simultaneously, they employ fully shared generators to extract features, resulting in the loss of domain-specific information and the destruction of performance. To address these issues, we propose a novel dual-attention-based adversarial network for motor imagery classification (MI-DABAN). Our framework leverages multiple subjects' knowledge to improve a single subject's motor imagery classification performance by cleverly using a novel adversarial learning method and two unshared attention blocks. Specifically, without introducing additional domain discriminators, we iteratively maximize and minimize the output difference between the two classifiers to implement adversarial learning to ensure accurate domain alignment. Among them, maximization is used to identify easily confused samples near the decision boundary, and minimization is used to align the source and target domain distributions. Moreover, for the shallow features from source and target domains, we use two non-shared attention blocks to preserve domain-specific information, which can prevent the negative transfer of domain information and further improve the classification performance on test data. We conduct extensive experiments on two publicly available EEG datasets, namely BCI Competition IV Datasets 2a and 2b. The experiment results demonstrate our method's effectiveness and superiority.

Dual attentive fusion for EEG-based brain-computer interfaces

The classification based on Electroencephalogram (EEG) is a challenging task in the brain-computer interface (BCI) field due to data with a low signal-to-noise ratio. Most current deep learning based studies in this challenge focus on designing a desired convolutional neural network (CNN) to learn and classify the raw EEG signals. However, only CNN itself may not capture the highly discriminative patterns of EEG due to a lack of exploration of attentive spatial and temporal dynamics. To improve information utilization, this study proposes a Dual Attentive Fusion Model (DAFM) for the EEG-based BCI. DAFM is employed to capture the spatial and temporal information by modeling the interdependencies between the features from the EEG signals. To our best knowledge, our method is the first to fuse spatial and temporal dimensions in an interactive attention module. This module improves the expression ability of the extracted features. Extensive experiments implemented on four publicly available datasets demonstrate that our method outperforms state-of-the-art methods. Meanwhile, this work also indicates the effectiveness of Dual Attentive Fusion Module.