Multi-scale convolutional transformer network for motor imagery brain-computer interface
Abstract Brain-computer interface (BCI) systems allow users to communicate with external devices by translating neural signals into real-time commands. Convolutional neural networks (CNNs) have been effectively utilized for decoding motor imagery electroencephalography (MI-EEG) signals in BCIs. Howe...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-04-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-96611-5 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850156513256013824 |
|---|---|
| author | Wei Zhao Baocan Zhang Haifeng Zhou Dezhi Wei Chenxi Huang Quan Lan |
| author_facet | Wei Zhao Baocan Zhang Haifeng Zhou Dezhi Wei Chenxi Huang Quan Lan |
| author_sort | Wei Zhao |
| collection | DOAJ |
| description | Abstract Brain-computer interface (BCI) systems allow users to communicate with external devices by translating neural signals into real-time commands. Convolutional neural networks (CNNs) have been effectively utilized for decoding motor imagery electroencephalography (MI-EEG) signals in BCIs. However, traditional CNN-based methods face challenges such as individual variability in EEG signals and the limited receptive fields of CNNs. This study presents the Multi-Scale Convolutional Transformer (MSCFormer) model that integrates multiple CNN branches for multi-scale feature extraction and a Transformer module to capture global dependencies, followed by a fully connected layer for classification. The multi-branch multi-scale CNN structure effectively addresses individual variability in EEG signals, enhancing the model’s generalization capabilities, while the Transformer encoder strengthens global feature integration and improves decoding performance. Extensive experiments on the BCI IV-2a and IV-2b datasets show that MSCFormer achieves average accuracies of 82.95% (BCI IV-2a) and 88.00% (BCI IV-2b), with kappa values of 0.7726 and 0.7599 in five-fold cross-validation, surpassing several state-of-the-art methods. These results highlight MSCFormer’s robustness and accuracy, underscoring its potential in EEG-based BCI applications. The code has been released in https://github.com/snailpt/MSCFormer . |
| format | Article |
| id | doaj-art-082dfb75f708451a93cf8ed179c1b70f |
| institution | OA Journals |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-082dfb75f708451a93cf8ed179c1b70f2025-08-20T02:24:30ZengNature PortfolioScientific Reports2045-23222025-04-0115111810.1038/s41598-025-96611-5Multi-scale convolutional transformer network for motor imagery brain-computer interfaceWei Zhao0Baocan Zhang1Haifeng Zhou2Dezhi Wei3Chenxi Huang4Quan Lan5Chengyi College, Jimei UniversityChengyi College, Jimei UniversitySchool of Marine Engineering, Jimei UniversityChengyi College, Jimei UniversitySchool of Informatics, Xiamen UniversityDepartment of Neurology, Department of Neuroscience, School of Medicine, The First Affiliated Hospital of Xiamen University, Xiamen UniversityAbstract Brain-computer interface (BCI) systems allow users to communicate with external devices by translating neural signals into real-time commands. Convolutional neural networks (CNNs) have been effectively utilized for decoding motor imagery electroencephalography (MI-EEG) signals in BCIs. However, traditional CNN-based methods face challenges such as individual variability in EEG signals and the limited receptive fields of CNNs. This study presents the Multi-Scale Convolutional Transformer (MSCFormer) model that integrates multiple CNN branches for multi-scale feature extraction and a Transformer module to capture global dependencies, followed by a fully connected layer for classification. The multi-branch multi-scale CNN structure effectively addresses individual variability in EEG signals, enhancing the model’s generalization capabilities, while the Transformer encoder strengthens global feature integration and improves decoding performance. Extensive experiments on the BCI IV-2a and IV-2b datasets show that MSCFormer achieves average accuracies of 82.95% (BCI IV-2a) and 88.00% (BCI IV-2b), with kappa values of 0.7726 and 0.7599 in five-fold cross-validation, surpassing several state-of-the-art methods. These results highlight MSCFormer’s robustness and accuracy, underscoring its potential in EEG-based BCI applications. The code has been released in https://github.com/snailpt/MSCFormer .https://doi.org/10.1038/s41598-025-96611-5Brain-computer interface (BCI)Convolutional neural networks (CNNs)Electroencephalography (EEG)Motor imagery (MI)Transformer |
| spellingShingle | Wei Zhao Baocan Zhang Haifeng Zhou Dezhi Wei Chenxi Huang Quan Lan Multi-scale convolutional transformer network for motor imagery brain-computer interface Scientific Reports Brain-computer interface (BCI) Convolutional neural networks (CNNs) Electroencephalography (EEG) Motor imagery (MI) Transformer |
| title | Multi-scale convolutional transformer network for motor imagery brain-computer interface |
| title_full | Multi-scale convolutional transformer network for motor imagery brain-computer interface |
| title_fullStr | Multi-scale convolutional transformer network for motor imagery brain-computer interface |
| title_full_unstemmed | Multi-scale convolutional transformer network for motor imagery brain-computer interface |
| title_short | Multi-scale convolutional transformer network for motor imagery brain-computer interface |
| title_sort | multi scale convolutional transformer network for motor imagery brain computer interface |
| topic | Brain-computer interface (BCI) Convolutional neural networks (CNNs) Electroencephalography (EEG) Motor imagery (MI) Transformer |
| url | https://doi.org/10.1038/s41598-025-96611-5 |
| work_keys_str_mv | AT weizhao multiscaleconvolutionaltransformernetworkformotorimagerybraincomputerinterface AT baocanzhang multiscaleconvolutionaltransformernetworkformotorimagerybraincomputerinterface AT haifengzhou multiscaleconvolutionaltransformernetworkformotorimagerybraincomputerinterface AT dezhiwei multiscaleconvolutionaltransformernetworkformotorimagerybraincomputerinterface AT chenxihuang multiscaleconvolutionaltransformernetworkformotorimagerybraincomputerinterface AT quanlan multiscaleconvolutionaltransformernetworkformotorimagerybraincomputerinterface |