Dual-Branch Convolution Network With Efficient Channel Attention for EEG-Based Motor Imagery Classification DOI Creative Commons
Kai Zhou,

Aierken Haimudula,

Wanying Tang

et al.

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 74930 - 74943

Published: Jan. 1, 2024

Brain-Computer Interface (BCI) is a revolutionary technique that employs wearable electroencephalography (EEG) sensors and artificial intelligence (AI) to monitor decode brain activity. EEG-based motor imagery (MI) signal widely utilized in various BCI fields including intelligent healthcare, robot control, smart homes. Yet, the limited capability of decoding signals remains significant obstacle techniques expansion. In this study, we describe an architecture known as dual-branch attention temporal convolutional network (DB-ATCNet) for MI classification. DB-ATCNet improves classification performance with relatively fewer parameters by utilizing channel attention. The model consists two primary modules: convolution (ADBC) fusion (ATFC). ADBC module utilizes extract low-level MI-EEG features incorporates improve spatial feature extraction. ATFC sliding windows self-attention obtain high-level features, strategies minimize information loss. achieved subject-independent accuracies 87.33% 69.58% two-class four-class tasks, respectively, on PhysioNet dataset. On Competition IV-2a dataset, it accuracy 71.34% 87.54% subject-dependent evaluations, surpassing existing methods. code available at https://github.com/zk-xju/DB-ATCNet.

Language: Английский

Multiclass classification of motor imagery tasks based on multi-branch convolutional neural network and temporal convolutional network model DOI
Shiqi Yu, Zedong Wang, Fei Wang

et al.

Cerebral Cortex, Journal Year: 2024, Volume and Issue: 34(2)

Published: Jan. 5, 2024

Abstract Motor imagery (MI) is a cognitive process wherein an individual mentally rehearses specific movement without physically executing it. Recently, MI-based brain–computer interface (BCI) has attracted widespread attention. However, accurate decoding of MI and understanding neural mechanisms still face huge challenges. These seriously hinder the clinical application development BCI systems based on MI. Thus, it very necessary to develop new methods decode tasks. In this work, we propose multi-branch convolutional network (MBCNN) with temporal (TCN), end-to-end deep learning framework multi-class We first used MBCNN capture electroencephalography signals information spectral domains through different kernels. Then, introduce TCN extract more discriminative features. The within-subject cross-session strategy validate classification performance dataset Competition IV-2a. results showed that achieved 75.08% average accuracy for 4-class task classification, outperforming several state-of-the-art approaches. proposed MBCNN-TCN-Net successfully captures features decodes tasks effectively, improving MI-BCIs. Our findings could provide significant potential systems.

Language: Английский

Citations

10
Multi-branch spatial-temporal-spectral convolutional neural networks for multi-task motor imagery EEG classification DOI

Zikun Cai,

Tian-jian Luo, Xuan Cao

et al.

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 93, P. 106156 - 106156

Published: Feb. 28, 2024

Language: Английский

Citations

9
Holographic convolutional attention neural network for motor imagery decoding based on EEG temporal–spatial frequency features DOI
Qingsong Ai,

Yuang Liu,

Quan Liu

et al.

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 104, P. 107526 - 107526

Published: Jan. 29, 2025

Language: Английский

Citations

1
Managing Distributed Machine Learning Lifecycle for Healthcare Data in the Cloud DOI Creative Commons
Engin Zeydan, Şuayb Ş. Arslan, Madhusanka Liyanage

et al.

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 115750 - 115774

Published: Jan. 1, 2024

Language: Английский

Citations

7
EEG-based motor imagery classification with quantum algorithms DOI
Cynthia Olvera, Oscar Montiel, Yoshio Rubio

et al.

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 247, P. 123354 - 123354

Published: Jan. 30, 2024

Language: Английский

Citations

6
A robust multi-branch multi-attention-mechanism EEGNet for motor imagery BCI decoding DOI
Haodong Deng, Mengfan Li,

Jundi Li

et al.

Journal of Neuroscience Methods, Journal Year: 2024, Volume and Issue: 405, P. 110108 - 110108

Published: March 6, 2024

Language: Английский

Citations

5
Multi-Task Heterogeneous Ensemble Learning-Based Cross-Subject EEG Classification Under Stroke Patients DOI Creative Commons
Minji Lee, Hyeong-Yeong Park, Wanjoo Park

et al.

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2024, Volume and Issue: 32, P. 1767 - 1778

Published: Jan. 1, 2024

Robot-assisted motor training is applied for neurorehabilitation in stroke patients, using imagery (MI) as a representative paradigm of brain-computer interfaces to offer real-life assistance individuals facing movement challenges. However, the effectiveness with MI may vary depending on location lesion, which should be considered. This paper introduces multi-task electroencephalogram-based heterogeneous ensemble learning (MEEG-HEL) specifically designed cross-subject training. In proposed framework, common spatial patterns were used feature extraction, and features according lesions are shared selected through sequential forward floating selection. The ensembles classifiers. Nine patients chronic ischemic participated, engaging execution (ME) paradigms involving finger tapping. classification criteria established two ways, taking into account characteristics patients. session, first involved direction recognition task two-handed classification, achieving performance 0.7419 (±0.0811) 0.7061 (±0.1270) ME. second focused assessment lesion location, resulting 0.7457 (±0.1317) 0.6791 (±0.1253) Comparing specific-subject except ME task, both tasks was significantly higher than session. Furthermore, similar or statistically sessions compared baseline models. MEEG-HEL holds promise improving practicality clinical settings facilitating detection lesions.

Language: Английский

Citations

5
EEG-FMCNN: A fusion multi-branch 1D convolutional neural network for EEG-based motor imagery classification DOI
Wenlong Wang, Baojiang Li, Haiyan Wang

et al.

Medical & Biological Engineering & Computing, Journal Year: 2023, Volume and Issue: 62(1), P. 107 - 120

Published: Sept. 20, 2023

Language: Английский

Citations

12
ETCNet: An EEG-based motor imagery classification model combining efficient channel attention and temporal convolutional network DOI
Yuxin Qin, Baojiang Li, Wenlong Wang

et al.

Brain Research, Journal Year: 2023, Volume and Issue: 1823, P. 148673 - 148673

Published: Nov. 11, 2023

Language: Английский

Citations

12
EISATC-Fusion: Inception Self-Attention Temporal Convolutional Network Fusion for Motor Imagery EEG Decoding DOI Creative Commons
Guangjin Liang, Dianguo Cao, Jinqiang Wang

et al.

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2024, Volume and Issue: 32, P. 1535 - 1545

Published: Jan. 1, 2024

The motor imagery brain-computer interface (MI-BCI) based on electroencephalography (EEG) is a widely used human-machine paradigm.However, due to the non-stationarity and individual differences among subjects in EEG signals, decoding accuracy limited, affecting application of MI-BCI.In this paper, we propose EISATC-Fusion model for MI decoding, consisting inception block, multi-head selfattention (MSA), temporal convolutional network (TCN), layer fusion.Specifically, design DS Inception block extract multi-scale frequency band information.And new cnnCosMSA module CNN cos attention solve collapse improve interpretability model.The TCN improved by depthwise separable convolution reduces parameters fusion consists feature decision fusion, fully utilizing features output enhances robustness model.We two-stage training strategy training.Early stopping prevent overfitting, loss validation set are as indicators early stopping.The proposed achieves within-subject classification accuracies 84.57% 87.58% BCI Competition IV Datasets 2a 2b, respectively.And cross-subject 67.42% 71.23% (by transfer learning) when with two sessions one session Dataset 2a, respectively.The demonstrated through weight visualization method.Index Terms-Brain-computer (BCI)

Language: Английский

Citations

4