Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer Interfaces

For practical motor imagery (MI) brain-computer interface (BCI) applications, generating a reliable model for a target subject with few MI trials is important since the data collection process is labour-intensive and expensive. In this paper, we address this issue by proposing a few-shot learning me...

Full description

Saved in:
Bibliographic Details
Main Authors: Xiuyu Huang, Shuang Liang, Yuanpeng Zhang, Nan Zhou, Witold Pedrycz, Kup-Sze Choi
Format: Article
Language:English
Published: IEEE 2023-01-01
Series:IEEE Transactions on Neural Systems and Rehabilitation Engineering
Subjects:
Online Access:https://ieeexplore.ieee.org/document/9978669/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849761984283672576
author Xiuyu Huang
Shuang Liang
Yuanpeng Zhang
Nan Zhou
Witold Pedrycz
Kup-Sze Choi
author_facet Xiuyu Huang
Shuang Liang
Yuanpeng Zhang
Nan Zhou
Witold Pedrycz
Kup-Sze Choi
author_sort Xiuyu Huang
collection DOAJ
description For practical motor imagery (MI) brain-computer interface (BCI) applications, generating a reliable model for a target subject with few MI trials is important since the data collection process is labour-intensive and expensive. In this paper, we address this issue by proposing a few-shot learning method called temporal episode relation learning (TERL). TERL models MI with only limited trials from the target subject by the ability to compare MI trials through episode-based training. It can be directly applied to a new user without being re-trained, which is vital to improve user experience and realize real-world MIBCI applications. We develop a new and effective approach where, unlike the original episode learning, the temporal pattern between trials in each episode is encoded during the learning to boost the classification performance. We also perform an online evaluation simulation, in addition to the offline analysis that the previous studies only conduct, to better understand the performance of different approaches in real-world scenario. Extensive experiments are completed on four publicly available MIBCI datasets to evaluate the proposed TERL. Results show that TERL outperforms baseline and recent state-of-the-art methods, demonstrating competitive performance for subject-specific MIBCI where few trials are available from a target subject and a considerable number of trials from other source subjects.
format Article
id doaj-art-71f105a4fee34aefa927be68d1e9b858
institution DOAJ
issn 1534-4320
1558-0210
language English
publishDate 2023-01-01
publisher IEEE
record_format Article
series IEEE Transactions on Neural Systems and Rehabilitation Engineering
spelling doaj-art-71f105a4fee34aefa927be68d1e9b8582025-08-20T03:05:52ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1534-43201558-02102023-01-013153054310.1109/TNSRE.2022.32282169978669Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer InterfacesXiuyu Huang0https://orcid.org/0000-0003-1600-9109Shuang Liang1https://orcid.org/0000-0003-0305-7558Yuanpeng Zhang2https://orcid.org/0000-0003-1736-3425Nan Zhou3https://orcid.org/0000-0002-0434-6231Witold Pedrycz4https://orcid.org/0000-0002-9335-9930Kup-Sze Choi5https://orcid.org/0000-0003-0836-7088Center for Smart Health, The Hong Kong Polytechnic University, Hong Kong, SAR, ChinaSmart Health Big Data Analysis and Location Services Engineering Laboratory of Jiangsu Province, Nanjing University of Posts and Telecommunications, Nanjing, ChinaMedical Informatics, Nantong University, Nantong, ChinaSchool of Electronic Information and Electronic Engineering, Chengdu University, Chengdu, ChinaDepartment of Electrical and Computer Engineering, University of Alberta, Edmonton, CanadaCenter for Smart Health, The Hong Kong Polytechnic University, Hong Kong, SAR, ChinaFor practical motor imagery (MI) brain-computer interface (BCI) applications, generating a reliable model for a target subject with few MI trials is important since the data collection process is labour-intensive and expensive. In this paper, we address this issue by proposing a few-shot learning method called temporal episode relation learning (TERL). TERL models MI with only limited trials from the target subject by the ability to compare MI trials through episode-based training. It can be directly applied to a new user without being re-trained, which is vital to improve user experience and realize real-world MIBCI applications. We develop a new and effective approach where, unlike the original episode learning, the temporal pattern between trials in each episode is encoded during the learning to boost the classification performance. We also perform an online evaluation simulation, in addition to the offline analysis that the previous studies only conduct, to better understand the performance of different approaches in real-world scenario. Extensive experiments are completed on four publicly available MIBCI datasets to evaluate the proposed TERL. Results show that TERL outperforms baseline and recent state-of-the-art methods, demonstrating competitive performance for subject-specific MIBCI where few trials are available from a target subject and a considerable number of trials from other source subjects.https://ieeexplore.ieee.org/document/9978669/Motor imagerybrain-computer interfacetemporal encodingepisode training
spellingShingle Xiuyu Huang
Shuang Liang
Yuanpeng Zhang
Nan Zhou
Witold Pedrycz
Kup-Sze Choi
Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer Interfaces
IEEE Transactions on Neural Systems and Rehabilitation Engineering
Motor imagery
brain-computer interface
temporal encoding
episode training
title Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer Interfaces
title_full Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer Interfaces
title_fullStr Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer Interfaces
title_full_unstemmed Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer Interfaces
title_short Relation Learning Using Temporal Episodes for Motor Imagery Brain-Computer Interfaces
title_sort relation learning using temporal episodes for motor imagery brain computer interfaces
topic Motor imagery
brain-computer interface
temporal encoding
episode training
url https://ieeexplore.ieee.org/document/9978669/
work_keys_str_mv AT xiuyuhuang relationlearningusingtemporalepisodesformotorimagerybraincomputerinterfaces
AT shuangliang relationlearningusingtemporalepisodesformotorimagerybraincomputerinterfaces
AT yuanpengzhang relationlearningusingtemporalepisodesformotorimagerybraincomputerinterfaces
AT nanzhou relationlearningusingtemporalepisodesformotorimagerybraincomputerinterfaces
AT witoldpedrycz relationlearningusingtemporalepisodesformotorimagerybraincomputerinterfaces
AT kupszechoi relationlearningusingtemporalepisodesformotorimagerybraincomputerinterfaces