Advancing BCI with a transformer-based model for motor imagery classification

Abstract Brain-computer interfaces (BCIs) harness electroencephalographic signals for direct neural control of devices, offering significant benefits for individuals with motor impairments. Traditional machine learning methods for EEG-based motor imagery (MI) classification encounter challenges such...

Full description

Saved in:
Bibliographic Details
Main Authors: Wangdan Liao, Hongyun Liu, Weidong Wang
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-06364-4
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849769335695867904
author Wangdan Liao
Hongyun Liu
Weidong Wang
author_facet Wangdan Liao
Hongyun Liu
Weidong Wang
author_sort Wangdan Liao
collection DOAJ
description Abstract Brain-computer interfaces (BCIs) harness electroencephalographic signals for direct neural control of devices, offering significant benefits for individuals with motor impairments. Traditional machine learning methods for EEG-based motor imagery (MI) classification encounter challenges such as manual feature extraction and susceptibility to noise. This paper introduces EEGEncoder, a deep learning framework that employs modified transformers and Temporal Convolutional Networks (TCNs) to surmount these limitations. We propose a novel fusion architecture, named Dual-Stream Temporal-Spatial Block (DSTS), to capture temporal and spatial features, improving the accuracy of Motor Imagery classification task. Additionally, we use multiple parallel structures to enhance the model’s performance. When tested on the BCI Competition IV-2a dataset, our proposed model achieved an average accuracy of 86.46% for subject dependent and average 74.48% for subject independent.
format Article
id doaj-art-ad091aafb6e643eba04d2f738898302e
institution DOAJ
issn 2045-2322
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-ad091aafb6e643eba04d2f738898302e2025-08-20T03:03:27ZengNature PortfolioScientific Reports2045-23222025-07-0115111210.1038/s41598-025-06364-4Advancing BCI with a transformer-based model for motor imagery classificationWangdan Liao0Hongyun Liu1Weidong Wang2School of Biological Science and Medical Engineering, Beihang UniversityMedical Innovation Research Division, Chinese PLA General HospitalSchool of Biological Science and Medical Engineering, Beihang UniversityAbstract Brain-computer interfaces (BCIs) harness electroencephalographic signals for direct neural control of devices, offering significant benefits for individuals with motor impairments. Traditional machine learning methods for EEG-based motor imagery (MI) classification encounter challenges such as manual feature extraction and susceptibility to noise. This paper introduces EEGEncoder, a deep learning framework that employs modified transformers and Temporal Convolutional Networks (TCNs) to surmount these limitations. We propose a novel fusion architecture, named Dual-Stream Temporal-Spatial Block (DSTS), to capture temporal and spatial features, improving the accuracy of Motor Imagery classification task. Additionally, we use multiple parallel structures to enhance the model’s performance. When tested on the BCI Competition IV-2a dataset, our proposed model achieved an average accuracy of 86.46% for subject dependent and average 74.48% for subject independent.https://doi.org/10.1038/s41598-025-06364-4Motor imagery (MI)Electroencephalography (EEG)ClassificationTransformerTemporal Convolutional Networks (TCNs)
spellingShingle Wangdan Liao
Hongyun Liu
Weidong Wang
Advancing BCI with a transformer-based model for motor imagery classification
Scientific Reports
Motor imagery (MI)
Electroencephalography (EEG)
Classification
Transformer
Temporal Convolutional Networks (TCNs)
title Advancing BCI with a transformer-based model for motor imagery classification
title_full Advancing BCI with a transformer-based model for motor imagery classification
title_fullStr Advancing BCI with a transformer-based model for motor imagery classification
title_full_unstemmed Advancing BCI with a transformer-based model for motor imagery classification
title_short Advancing BCI with a transformer-based model for motor imagery classification
title_sort advancing bci with a transformer based model for motor imagery classification
topic Motor imagery (MI)
Electroencephalography (EEG)
Classification
Transformer
Temporal Convolutional Networks (TCNs)
url https://doi.org/10.1038/s41598-025-06364-4
work_keys_str_mv AT wangdanliao advancingbciwithatransformerbasedmodelformotorimageryclassification
AT hongyunliu advancingbciwithatransformerbasedmodelformotorimageryclassification
AT weidongwang advancingbciwithatransformerbasedmodelformotorimageryclassification