Dual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification

Event-Related Desynchronization (ERD) or Electroencephalogram (EEG) wavelet is essential for motor imagery (MI) classification and BMI (Brain–Machine Interface) application. However, it is difficult to recognize multiple tasks for non-trained subjects that are indispensable for the complexities of t...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Meiyan Xu, Junfeng Yao, Hualiang Ni
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	machine learning brain machine interface attention motor imagery classification Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
Acceso en línea:	https://doaj.org/article/cafd732462fd4b6ab1780d2922d7139b
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:cafd732462fd4b6ab1780d2922d7139b
record_format	dspace
spelling	oai:doaj.org-article:cafd732462fd4b6ab1780d2922d7139b2021-11-25T16:40:48ZDual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification10.3390/app1122109062076-3417https://doaj.org/article/cafd732462fd4b6ab1780d2922d7139b2021-11-01T00:00:00Zhttps://www.mdpi.com/2076-3417/11/22/10906https://doaj.org/toc/2076-3417Event-Related Desynchronization (ERD) or Electroencephalogram (EEG) wavelet is essential for motor imagery (MI) classification and BMI (Brain–Machine Interface) application. However, it is difficult to recognize multiple tasks for non-trained subjects that are indispensable for the complexities of the task or the uncertainties in the environment. The subject-independent scenario, where an inter-subject trained model can be directly applied to new users without precalibration, is particularly desired. Therefore, this paper focuses on an effective attention mechanism which can be applied to a subject-independent set to learn EEG motor imagery features. Firstly, a custom form of sequence inputs with spatial and temporal dimensions is adopted for dual headed attention via deep convolution net (DHDANet). Secondly, DHDANet simultaneously learns temporal and spacial features. The features of spacial attention on each input head are divided into two parts for spatial attentional learning subsequently. The proposed model is validated based on the EEG-MI signals collected from 54 subjects in two sessions with 200 trials in each sessions. The classification of left and right hand motor imagery in this paper achieves an average accuracy of 75.52%, a significant improvement compared to state-of-the-art methods. In addition, the visualization of the frequency analysis method demonstrates that the temporal-convolution and spectral-attention is capable of identifying the ERD for EEG-MI. The proposed machine learning structure enables cross-session and cross-subject classification and makes significant progress in the BMI transfer learning problem.Meiyan XuJunfeng YaoHualiang NiMDPI AGarticlemachine learningbrain machine interfaceattentionmotor imageryclassificationTechnologyTEngineering (General). Civil engineering (General)TA1-2040Biology (General)QH301-705.5PhysicsQC1-999ChemistryQD1-999ENApplied Sciences, Vol 11, Iss 10906, p 10906 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	machine learning brain machine interface attention motor imagery classification Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
spellingShingle	machine learning brain machine interface attention motor imagery classification Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999 Meiyan Xu Junfeng Yao Hualiang Ni Dual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification
description	Event-Related Desynchronization (ERD) or Electroencephalogram (EEG) wavelet is essential for motor imagery (MI) classification and BMI (Brain–Machine Interface) application. However, it is difficult to recognize multiple tasks for non-trained subjects that are indispensable for the complexities of the task or the uncertainties in the environment. The subject-independent scenario, where an inter-subject trained model can be directly applied to new users without precalibration, is particularly desired. Therefore, this paper focuses on an effective attention mechanism which can be applied to a subject-independent set to learn EEG motor imagery features. Firstly, a custom form of sequence inputs with spatial and temporal dimensions is adopted for dual headed attention via deep convolution net (DHDANet). Secondly, DHDANet simultaneously learns temporal and spacial features. The features of spacial attention on each input head are divided into two parts for spatial attentional learning subsequently. The proposed model is validated based on the EEG-MI signals collected from 54 subjects in two sessions with 200 trials in each sessions. The classification of left and right hand motor imagery in this paper achieves an average accuracy of 75.52%, a significant improvement compared to state-of-the-art methods. In addition, the visualization of the frequency analysis method demonstrates that the temporal-convolution and spectral-attention is capable of identifying the ERD for EEG-MI. The proposed machine learning structure enables cross-session and cross-subject classification and makes significant progress in the BMI transfer learning problem.
format	article
author	Meiyan Xu Junfeng Yao Hualiang Ni
author_facet	Meiyan Xu Junfeng Yao Hualiang Ni
author_sort	Meiyan Xu
title	Dual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification
title_short	Dual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification
title_full	Dual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification
title_fullStr	Dual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification
title_full_unstemmed	Dual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification
title_sort	dual head and dual attention in deep learning for end-to-end eeg motor imagery classification
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/cafd732462fd4b6ab1780d2922d7139b
work_keys_str_mv	AT meiyanxu dualheadanddualattentionindeeplearningforendtoendeegmotorimageryclassification AT junfengyao dualheadanddualattentionindeeplearningforendtoendeegmotorimageryclassification AT hualiangni dualheadanddualattentionindeeplearningforendtoendeegmotorimageryclassification
_version_	1718413102854176768

Dual Head and Dual Attention in Deep Learning for End-to-End EEG Motor Imagery Classification

Ejemplares similares