EEGProgress: A fast and lightweight progressive convolution architecture for EEG classification

Computers in Biology and Medicine, Год журнала: 2023, Номер 169, С. 107901 - 107901

Опубликована: Дек. 27, 2023

Язык: Английский

---

A multiscale convolutional neural network based on time-frequency features for decoding rat exercise fatigue LFP

Biomedical Signal Processing and Control, Год журнала: 2024, Номер 95, С. 106333 - 106333

Опубликована: Апрель 18, 2024

Язык: Английский

---

Optimal Channel Selection of Multiclass Motor Imagery Classification Based on Fusion Convolutional Neural Network with Attention Blocks

Sensors, Год журнала: 2024, Номер 24(10), С. 3168 - 3168

Опубликована: Май 16, 2024

The widely adopted paradigm in brain–computer interfaces (BCIs) involves motor imagery (MI), enabling improved communication between humans and machines. EEG signals derived from MI present several challenges due to their inherent characteristics, which lead a complex process of classifying finding the potential tasks specific participant. Another issue is that BCI systems can result noisy data redundant channels, turn increased equipment computational costs. To address these problems, optimal channel selection multiclass classification based on Fusion convolutional neural network with Attention blocks (FCNNA) proposed. In this study, we developed CNN model consisting layers multiple spatial temporal filters. These filters are designed specifically capture distribution relationships signal features across different electrode locations, as well analyze evolution over time. Following layers, Convolutional Block Module (CBAM) used to, further, enhance feature extraction. selection, genetic algorithm select set channels using new technique deliver fixed variable for all participants. proposed methodology validated showing 6.41% improvement compared most baseline models. Notably, achieved highest results 93.09% binary classes involving left-hand right-hand movements. addition, cross-subject strategy yielded an impressive accuracy 68.87%. was enhanced, reaching 84.53%. Overall, our experiments illustrated efficiency both classification, superior either full or reduced number channels.

Язык: Английский

---

Motor Imagery EEG Signal Classification Using Distinctive Feature Fusion with Adaptive Structural LASSO

Sensors, Год журнала: 2024, Номер 24(12), С. 3755 - 3755

Опубликована: Июнь 9, 2024

A motor imagery brain–computer interface connects the human brain and computers via electroencephalography (EEG). However, individual differences in frequency ranges of activity during tasks pose a challenge, limiting manual feature extraction for classification. To extract features that match specific subjects, we proposed novel classification model using distinctive fusion with adaptive structural LASSO. Specifically, extracted spatial domain from overlapping multi-scale sub-bands EEG signals mined discriminative by fusing task relevance information into LASSO-based selection. We evaluated on public datasets, demonstrating has excellent performance. Meanwhile, ablation studies selection visualization further verified great potential analysis.

Язык: Английский

---

MI-MBFT: Superior Motor Imagery Decoding of Raw EEG Data Based on a Multi-Branch and Fusion Transformer Framework

IEEE Sensors Journal, Год журнала: 2024, Номер 24(21), С. 34879 - 34891

Опубликована: Авг. 28, 2024

Язык: Английский

---

A parallel-hierarchical neural network (PHNN) for motor imagery EEG signal classification

Biomedical Signal Processing and Control, Год журнала: 2023, Номер 88, С. 105621 - 105621

Опубликована: Окт. 20, 2023

Язык: Английский

---

Efficient Feature Learning Model of Motor Imagery EEG Signals with L1-Norm and Weighted Fusion

Biosensors, Год журнала: 2024, Номер 14(5), С. 211 - 211

Опубликована: Апрель 23, 2024

Brain–computer interface (BCI) for motor imagery is an advanced technology used in the field of medical rehabilitation. However, due to poor accuracy electroencephalogram feature classification, BCI systems often misrecognize user commands. Although many state-of-the-art selection methods aim enhance classification accuracy, they usually overlook interrelationships between individual features, indirectly impacting classification. To overcome this issue, we propose adaptive learning model that employs a Riemannian geometric approach generate matrix from signals, serving as model’s input. By integrating enhanced L1 penalty and weighted fusion into sparse model, select most informative features matrix. Specifically, measure importance using mutual information introduce weight construction strategy penalize regression coefficients corresponding each variable adaptively. Moreover, balances differences among correlated variables, reducing overreliance on specific variables enhancing accuracy. The performance proposed method was validated Competition IV datasets IIa IIb support vector machine. Experimental results demonstrate effectiveness superiority compared existing models.

Язык: Английский

---

Effectiveness of Adaptive Attention-based Network for Situation Awareness Recognition

IEEE Sensors Journal, Год журнала: 2024, Номер 24(12), С. 20092 - 20102

Опубликована: Май 8, 2024

Situation awareness (SA) is directly related to the operating level of dynamic system operators, and electroencephalography (EEG) frequently employed as gold standard for SA recognition. Several deep learning models performed well in recognition based on EEG features. However, it remains limitations such a limited size datasets, restricted model interpretability, low capability extracting beneficial In this work, an adaptive spatial-channel attention mechanism (ASCAM) was introduced architectures convolutional neural network (CNN). Specifically, ASCAM allows layers CNN fuse various sizes received information selectively focus effective interpretable Regarding problem by combing frequency noise with multivariate variational mode decomposition (MVMD) enhance generalization models. Experiment results gave that EEGNet embedded framework exhibited relative improvement 6.02% over baseline method. And contributes feature extraction significantly enhances considerable performance. Ablation studies were further implemented confirm efficacy proposed MVMD-based data augmentation. Interpretation indicated have discovered neurobiological mechanisms loss. Meanwhile, lightweight plug-and-play, which can be into any architecture utilized decoding tasks.

Язык: Английский

---

An intelligent mangosteen grading system based on an improved convolutional neural network

Signal Image and Video Processing, Год журнала: 2024, Номер 18(12), С. 8585 - 8595

Опубликована: Авг. 9, 2024

Язык: Английский

---

CWSTR-Net: A Channel-Weighted Spatial–Temporal Residual Network based on nonsmooth nonnegative matrix factorization for fatigue detection using EEG signals

Biomedical Signal Processing and Control, Год журнала: 2024, Номер 97, С. 106685 - 106685

Опубликована: Авг. 9, 2024

Язык: Английский

---