TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype Classification

Electroencephalogram (EEG) based seizure subtype classification is very important in clinical diagnostics. However, manual seizure subtype classification is expensive and time-consuming, whereas automatic classification usually needs a large number of labeled samples for model training. This paper p...

Full description

Saved in:
Bibliographic Details
Main Authors: Ruimin Peng, Changming Zhao, Jun Jiang, Guangtao Kuang, Yuqi Cui, Yifan Xu, Hao Du, Jianbo Shao, Dongrui Wu
Format: Article
Language:English
Published: IEEE 2022-01-01
Series:IEEE Transactions on Neural Systems and Rehabilitation Engineering
Subjects:
Online Access:https://ieeexplore.ieee.org/document/9878131/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849762758014271488
author Ruimin Peng
Changming Zhao
Jun Jiang
Guangtao Kuang
Yuqi Cui
Yifan Xu
Hao Du
Jianbo Shao
Dongrui Wu
author_facet Ruimin Peng
Changming Zhao
Jun Jiang
Guangtao Kuang
Yuqi Cui
Yifan Xu
Hao Du
Jianbo Shao
Dongrui Wu
author_sort Ruimin Peng
collection DOAJ
description Electroencephalogram (EEG) based seizure subtype classification is very important in clinical diagnostics. However, manual seizure subtype classification is expensive and time-consuming, whereas automatic classification usually needs a large number of labeled samples for model training. This paper proposes an EEGNet-based slim deep neural network, which relieves the labeled data requirement in EEG-based seizure subtype classification. A temporal information enhancement module with sinusoidal encoding is used to augment the first convolution layer of EEGNet. A training strategy for automatic hyper-parameter selection is also proposed. Experiments on the public TUSZ dataset and our own CHSZ dataset with infants and children demonstrated that our proposed TIE-EEGNet outperformed several traditional and deep learning models in cross-subject seizure subtype classification. Additionally, it also achieved the best performance in a challenging transfer learning scenario. Both our code and the CHSZ dataset are publicized.
format Article
id doaj-art-2c92b4dca96b4f42b0d87acc9df9f77c
institution DOAJ
issn 1534-4320
1558-0210
language English
publishDate 2022-01-01
publisher IEEE
record_format Article
series IEEE Transactions on Neural Systems and Rehabilitation Engineering
spelling doaj-art-2c92b4dca96b4f42b0d87acc9df9f77c2025-08-20T03:05:39ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1534-43201558-02102022-01-01302567257610.1109/TNSRE.2022.32045409878131TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype ClassificationRuimin Peng0https://orcid.org/0000-0002-4869-1328Changming Zhao1Jun Jiang2Guangtao Kuang3Yuqi Cui4Yifan Xu5https://orcid.org/0000-0003-1591-0384Hao Du6Jianbo Shao7Dongrui Wu8https://orcid.org/0000-0002-7153-9703Key Laboratory of theMinistry of Education for Image Processing and Intelligent Control, School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, ChinaKey Laboratory of theMinistry of Education for Image Processing and Intelligent Control, School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, ChinaChildren’s Hospital, Wuhan, ChinaChildren’s Hospital, Wuhan, ChinaKey Laboratory of theMinistry of Education for Image Processing and Intelligent Control, School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, ChinaKey Laboratory of theMinistry of Education for Image Processing and Intelligent Control, School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, ChinaChildren’s Hospital, Wuhan, ChinaChildren’s Hospital, Wuhan, ChinaKey Laboratory of theMinistry of Education for Image Processing and Intelligent Control, School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, ChinaElectroencephalogram (EEG) based seizure subtype classification is very important in clinical diagnostics. However, manual seizure subtype classification is expensive and time-consuming, whereas automatic classification usually needs a large number of labeled samples for model training. This paper proposes an EEGNet-based slim deep neural network, which relieves the labeled data requirement in EEG-based seizure subtype classification. A temporal information enhancement module with sinusoidal encoding is used to augment the first convolution layer of EEGNet. A training strategy for automatic hyper-parameter selection is also proposed. Experiments on the public TUSZ dataset and our own CHSZ dataset with infants and children demonstrated that our proposed TIE-EEGNet outperformed several traditional and deep learning models in cross-subject seizure subtype classification. Additionally, it also achieved the best performance in a challenging transfer learning scenario. Both our code and the CHSZ dataset are publicized.https://ieeexplore.ieee.org/document/9878131/EEGseizure subtype classificationEEGNetdeep learningpositional encoding
spellingShingle Ruimin Peng
Changming Zhao
Jun Jiang
Guangtao Kuang
Yuqi Cui
Yifan Xu
Hao Du
Jianbo Shao
Dongrui Wu
TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype Classification
IEEE Transactions on Neural Systems and Rehabilitation Engineering
EEG
seizure subtype classification
EEGNet
deep learning
positional encoding
title TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype Classification
title_full TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype Classification
title_fullStr TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype Classification
title_full_unstemmed TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype Classification
title_short TIE-EEGNet: Temporal Information Enhanced EEGNet for Seizure Subtype Classification
title_sort tie eegnet temporal information enhanced eegnet for seizure subtype classification
topic EEG
seizure subtype classification
EEGNet
deep learning
positional encoding
url https://ieeexplore.ieee.org/document/9878131/
work_keys_str_mv AT ruiminpeng tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification
AT changmingzhao tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification
AT junjiang tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification
AT guangtaokuang tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification
AT yuqicui tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification
AT yifanxu tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification
AT haodu tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification
AT jianboshao tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification
AT dongruiwu tieeegnettemporalinformationenhancedeegnetforseizuresubtypeclassification