Explore the vast range of titles available.

Transformer, a deep learning model with the self-attention mechanism, combined with the convolution neural network (CNN) has been successfully applied for decoding electroencephalogram (EEG) signals in Motor Imagery (MI) Brain-Computer Interface (BCI). However, the extremely non-linear, nonstationary characteristics of the EEG signals limits the effectiveness and efficiency of the deep learning methods. In addition, the variety of subjects and the experimental sessions impact the model adaptability. In this study, we propose a local and global convolutional transformer-based approach for MI-EEG classification. The local transformer encoder is combined to dynamically extract temporal features and make up for the shortcomings of the CNN model. The spatial features from all channels and the difference in hemispheres are obtained to improve the robustness of the model. To acquire adequate temporal-spatial feature representations, we combine the global transformer encoder and Densely Connected Network to improve the information flow and reuse. To validate the performance of the proposed model, three scenarios including within-session, cross-session and two-session are designed. In the experiments, the proposed method achieves up to 1.46%, 7.49% and 7.46% accuracy improvement respectively in the three scenarios for the public Korean dataset compared with current state-of-the-art models. For the BCI competition IV 2a dataset, the proposed model also achieves a 2.12% and 2.21% improvement for the cross-session and two-session scenarios respectively. The results confirm that the proposed approach can effectively extract much richer set of MI features from the EEG signals and improve the performance in the BCI applications.

IntroductionApproximately half of stroke survivors experience persistent upper limb dysfunction, which impairs self-care, reduces independence and lowers quality of life. Electroacupuncture is an established intervention with evidence supporting its role in improving upper limb motor function following ischaemic stroke. Motor imagery training (MIT), which activates the sensorimotor cortex through the mental rehearsal of movement, has shown promise as an adjunctive therapy in stroke rehabilitation. The concurrent application of electroacupuncture and MIT may enhance sensorimotor recovery by promoting the integration of central and peripheral neural pathways, potentially establishing a central–peripheral–central closed-loop circuit. However, empirical evidence supporting this integrative approach remains limited.This study aims to investigate the effects of electroacupuncture synchronised with MIT on upper limb function in patients with ischaemic stroke. In addition, longitudinal analysis of multimodal neuroimaging data will be used to explore the associated neural mechanisms.Methods and analysisA total of 72 patients with ischaemic stroke will be enrolled and randomly assigned (1:1) to receive either electroacupuncture synchronised with MIT or electroacupuncture. Each group will undergo 20 treatment sessions over 4 weeks (5 times per week). All participants will also receive standardised conventional rehabilitation training.The primary outcome is the Fugl-Meyer Assessment for the upper extremity. Secondary outcomes include the Modified Barthel Index for activities of daily living, the Modified Ashworth Scale (MAS) for spasticity, Brunnstrom stages, the 17-item Hamilton Depression Rating Scale, the Chinese version of the Massachusetts Acupuncture Sensation Scale and the Kinaesthetic and Visual Imagery Questionnaire. Assessments will be conducted at baseline, mid-treatment, post-treatment and at 8-week follow-up. In addition, functional connectivity of the cerebral cortex will be assessed using functional near-infrared spectroscopy and electroencephalography, which may serve as potential biomarkers of treatment response.Ethics and disseminationThis study has been approved by the Ethics Committee of Shanghai Second Rehabilitation Hospital (approval number: 2025-18-01) and has been registered with the International Traditional Medicine Clinical Trial Registry (ITMCTR; registration number: ITMCTR2025001311). The study will be conducted in accordance with the Declaration of Helsinki, relevant local regulations and applicable clinical guidelines. Informed consent will be obtained from all participants or their legal guardians, where applicable. The results will be disseminated through peer-reviewed publications and presentations at scientific conferences.Trial registration numberITMCTR2025001311.

Transfer learning enables the adaption of models to handle mismatches of distributions across sessions or across subjects. In this paper, we proposed a new transfer learning algorithm to classify motor imagery EEG data. By analyzing the power spectrum of EEG data related to motor imagery, the shared features across sessions or across subjects, namely, the mean and variance of model parameters, are extracted. Then, select the data sets that were most relevant to the new data set according to Euclidean distance to update the shared features. Finally, utilize the shared features and subject/session-specific features jointly to generate a new model. We evaluated our algorithm by analyzing the motor imagery EEG data from 10 healthy participants and a public data set from BCI competition IV. The classification accuracy of the proposed transfer learning is higher than that of traditional machine learning algorithms. The results of the paired t test showed that the classification results of PSD and the transfer learning algorithm were significantly different (p = 2.0946e-9), and the classification results of CSP and the transfer learning algorithm were significantly different (p = 1.9122e-6). The test accuracy of data set 2a of BCI competition IV was 85.7% ± 5.4%, which was higher than that of related traditional machine learning algorithms. Preliminary results suggested that the proposed algorithm can be effectively applied to the classification of motor imagery EEG signals across sessions and across subjects and the performance is better than that of the traditional machine learning algorithms. It can be promising to be applied to the field of brain-computer interface (BCI).

Background Growing evidence indicates a link between insulin resistance and post-stroke depression (PSD). This study employed the triglyceride glucose (TyG) index as a measure of insulin resistance to investigate its relationship with PSD. Methods This cross-sectional study utilized data from the National Health and Nutrition Examination Survey (2005–2018). PSD was assessed using data from patient health questionnaires, while the TyG index was calculated based on fasting venous blood glucose and fasting triglyceride levels. The formula used for the TyG index is ln[triglycerides (mg/dL) × fasting blood glucose (mg/dL)/2]. Participants were categorized into four groups according to the TyG index quartiles. A weighted multivariable logistic regression model was applied to examine the relationship between the TyG index and PSD. Results A total of 1217 patients were included in the study, of which 232 were diagnosed with PSD. The TyG index was divided into quartiles (Q1-Q4) for analysis. After adjusting for potential confounders, we found a significant positive association between the highest quartile of the TyG index (Q4: ≥9.33) and PSD (OR = 2.51, 95% CI: 1.04–6.07, p  = 0.041). This suggests that in the U.S. adult stroke population, individuals with higher TyG indices are more likely to experience depressive symptoms. Subgroup analysis further confirmed a stable and independent positive association between the TyG index and PSD (all trend p  > 0.05). Conclusion In this large cross-sectional study, our results suggest that among US adults who have experienced a stroke, those with higher TyG index levels are more likely to exhibit depressive symptoms. This provides a novel approach for the clinical prevention of PSD. Patients with higher TyG indices in the stroke population may require closer psychological health monitoring and timely intervention. Additionally, since the TyG index is calculated using only fasting blood glucose and triglyceride levels, it can help identify high-risk PSD patients, particularly in regions with limited healthcare resources.

In building a practical and robust brain-computer interface (BCI), the classification of motor imagery (MI) from electroencephalography (EEG) across multiple days is a long-standing challenge due to the large variability of the EEG signals. We collected a large dataset of MI from 5 different days with 25 subjects, the first open-access dataset to address BCI issues across 5 different days with a large number of subjects. The dataset includes 5 session data from 5 different days (2–3 days apart) for each subject. Each session contains 100 trials of left-hand and right-hand MI. In this report, we provide the benchmarking classification accuracy for three conditions, namely, within-session classification (WS), cross-session classification (CS), and cross-session adaptation (CSA), with subject-specific models. WS achieves an average classification accuracy of up to 68.8%, while CS degrades the accuracy to 53.7% due to the cross-session variability. However, by adaptation, CSA improves the accuracy to 78.9%. We anticipate this new dataset will significantly push further progress in MI BCI research in addressing the cross-session and cross-subject challenge.Measurement(s)ElectroencephalographyTechnology Type(s)motor imagery

Recently, convolutional neural networks (CNNs) have been widely applied in brain-computer interface (BCI) based on electroencephalogram (EEG) signals. Due to the subject-specific nature of EEG signal patterns and the multi-dimensionality of EEG features, it is necessary to employ appropriate feature representation methods to enhance the decoding accuracy of EEG. In this study, we proposed a method for representing EEG temporal, frequency, and phase features, aiming to preserve the multi-domain information of EEG signals. Specifically, we generated EEG temporal segments using a sliding window strategy. Then, temporal, frequency, and phase features were extracted from different temporal segments and stacked into 3D feature maps, namely temporal-frequency-phase features (TFPF). Furthermore, we designed a compact 3D-CNN model to extract these multi-domain features efficiently. Considering the inter-individual variability in EEG data, we conducted individual testing for each subject. The proposed model achieved an average accuracy of 89.86, 78.85, and 63.55% for 2-class, 3-class, and 4-class motor imagery (MI) classification tasks, respectively, on the PhysioNet dataset. On the GigaDB dataset, the average accuracy for 2-class MI classification was 91.91%. For the comparison between MI and real movement (ME) tasks, the average accuracy for the 2-class were 87.66 and 80.13% on the PhysioNet and GigaDB datasets, respectively. Overall, the method presented in this paper have obtained good results in MI/ME tasks and have a good application prospect in the development of BCI systems based on MI/ME.

The utilization of motor imagery-based brain-computer interfaces (MI-BCI) has been shown to assist stroke patients activate motor regions in the brain. In particular, the brain regions activated by unilateral upper limb multi-task are more extensive, which is more beneficial for rehabilitation, but it also increases the difficulty of decoding. In this paper, self-attention convolutional neural network based partial prior transfer learning (SACNN-PPTL) is proposed to improve the classification performance of patients’ MI multi-task. The backbone network of the algorithm is SACNN, which accords with the inherent features of electroencephalogram (EEG) and contains the temporal feature module, the spatial feature module and the feature generalization module. In addition, PPTL is introduced to transfer part of the target domain while preserving the generalization of the base model while improving the specificity of the target domain. In the experiment, five backbone networks and three training modes are selected as comparison algorithms. The experimental results show that SACNN-PPTL had a classification accuracy of 55.4%±0.17 in four types of MI tasks for 22 patients, which is significantly higher than comparison algorithms ( P  < 0.05). SACNN-PPTL effectively improves the decoding performance of MI tasks and promotes the development of BCI-based rehabilitation for unilateral upper limb.

In 2019, the International Classification of Diseases 11th Revision International Classification of Diseases (ICD-11) put forward a new concept of “chronic primary pain” (CPP), a kind of chronic pain characterized by severe functional disability and emotional distress, which is a medical problem that deserves great attention. Although CPP is closely related to depressive disorder, its potential neural characteristics are still unclear. This paper collected EEG data from 67 subjects (23 healthy subjects, 22 patients with depression, and 22 patients with CPP) under the auditory oddball paradigm, systematically analyzed the brain network connection matrix and graph theory characteristic indicators, and classified the EEG and PLI matrices of three groups of people by frequency band based on deep learning. The results showed significant differences in brain network connectivity between CPP patients and depressive patients. Specifically, the connectivity within the frontoparietal network of the Theta band in CPP patients is significantly enhanced. The CNN classification model of EEG is better than that of PLI, with the highest accuracy of 85.01% in Gamma band in former and 79.64% in Theta band in later. We propose hyperexcitability in attentional control in CPP patients and provide a novel method for objective assessment of chronic primary pain.

Background: The adoption of convolutional neural networks (CNNs) for decoding electroencephalogram (EEG)-based motor imagery (MI) in brain-computer interfaces has significantly increased recently. The effective extraction of motor imagery features is vital due to the variability among individuals and temporal states. Methods: This study introduces a novel network architecture, 3D-convolutional neural network-generative adversarial network (3D-CNN-GAN), for decoding both within-session and cross-session motor imagery. Initially, EEG signals were extracted over various time intervals using a sliding window technique, capturing temporal, frequency, and phase features to construct a temporal-frequency-phase feature (TFPF) three-dimensional feature map. Generative adversarial networks (GANs) were then employed to synthesize artificial data, which, when combined with the original datasets, expanded the data capacity and enhanced functional connectivity. Moreover, GANs proved capable of learning and amplifying the brain connectivity patterns present in the existing data, generating more distinctive brain network features. A compact, two-layer 3D-CNN model was subsequently developed to efficiently decode these TFPF features. Results: Taking into account session and individual differences in EEG data, tests were conducted on both the public GigaDB dataset and the SHU laboratory dataset. On the GigaDB dataset, our 3D-CNN and 3D-CNN-GAN models achieved two-class within-session motor imagery accuracies of 76.49% and 77.03%, respectively, demonstrating the algorithm’s effectiveness and the improvement provided by data augmentation. Furthermore, on the SHU dataset, the 3D-CNN and 3D-CNN-GAN models yielded two-class within-session motor imagery accuracies of 67.64% and 71.63%, and cross-session motor imagery accuracies of 58.06% and 63.04%, respectively. Conclusions: The 3D-CNN-GAN algorithm significantly enhances the generalizability of EEG-based motor imagery brain-computer interfaces (BCIs). Additionally, this research offers valuable insights into the potential applications of motor imagery BCIs.

A key challenge in developing a robust electroencephalography (EEG)-based brain-computer interface (BCI) is obtaining reliable classification performance across multiple days. In particular, EEG-based motor imagery (MI) BCI faces large variability and low signal-to-noise ratio. To address these issues, collecting a large and reliable dataset is critical for learning of cross-session and cross-subject patterns while mitigating EEG signals inherent instability. In this study, we obtained a comprehensive MI dataset from the 2019 World Robot Conference Contest-BCI Robot Contest. We collected EEG data from 62 healthy participants across three recording sessions. This experiment includes two paradigms: (1) two-class tasks: left and right hand-grasping, (2) three-class tasks: left and right hand-grasping, and foot-hooking. The dataset comprises raw data, and preprocessed data. For the two-class data, an average classification accuracy of 85.32% was achieved using EEGNet, while the three-class data achieved an accuracy of 76.90% using deepConvNet. Different researchers can reuse the dataset according to their needs. We hope that this dataset will significantly advance MI-BCI research, particularly in addressing cross-session and cross-subject challenges.