Epileptic Seizure Detection by Cascading Isolation Forest-Based Anomaly Screening and EasyEnsemble
The electroencephalogram (EEG), for measuring the electrophysiological activity of the brain, has been widely applied in automatic detection of epilepsy seizures. Various EEG-based seizure detection algorithms have already yielded high sensitivity, but training those algorithms requires a large amou...
Main Authors: | , , , , , , , , , , , |
---|---|
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/9745094/ |
_version_ | 1797805168931110912 |
---|---|
author | Yao Guo Xinyu Jiang Linkai Tao Long Meng Chenyun Dai Xi Long Feng Wan Yuan Zhang Johannes van Dijk Ronald M. Aarts Wei Chen Chen Chen |
author_facet | Yao Guo Xinyu Jiang Linkai Tao Long Meng Chenyun Dai Xi Long Feng Wan Yuan Zhang Johannes van Dijk Ronald M. Aarts Wei Chen Chen Chen |
author_sort | Yao Guo |
collection | DOAJ |
description | The electroencephalogram (EEG), for measuring the electrophysiological activity of the brain, has been widely applied in automatic detection of epilepsy seizures. Various EEG-based seizure detection algorithms have already yielded high sensitivity, but training those algorithms requires a large amount of labelled data. Data labelling is often done with a lot of human efforts, which is very time-consuming. In this study, we propose a hybrid system integrating an unsupervised learning (UL) module and a supervised learning (SL) module, where the UL module can significantly reduce the workload of data labelling. For preliminary seizure screening, UL synthesizes amplitude-integrated EEG (aEEG) extraction, isolation forest-based anomaly detection, adaptive segmentation, and silhouette coefficient-based anomaly detection evaluation. The UL module serves to quickly locate the determinate subjects (seizure segments and seizure-free segments) and the indeterminate subjects (potential seizure candidates). Afterwards, more robust seizure detection for the indeterminate subjects is performed by the SL using an EasyEnsemble algorithm. EasyEnsemble, as a class-imbalance learning method, can potentially decrease the generalization error of the seizure-free segments. The proposed method can significantly reduce the workload of data labelling while guaranteeing satisfactory performance. The proposed seizure detection system is evaluated using the Children’s Hospital Boston-Massachusetts Institute of Technology (CHB-MIT) scalp EEG dataset, and it achieves a mean accuracy of 92.62%, a mean sensitivity of 95.55%, and a mean specificity of 92.57%. To the best of our knowledge, this is the first epilepsy seizure detection study employing the integration of both the UL and the SL modules, achieving a competitive performance superior or similar to that of the state-of-the-art methods. |
first_indexed | 2024-03-13T05:47:59Z |
format | Article |
id | doaj.art-9634d59f63bd4824af9f6766d0de3cfb |
institution | Directory Open Access Journal |
issn | 1558-0210 |
language | English |
last_indexed | 2024-03-13T05:47:59Z |
publishDate | 2022-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
spelling | doaj.art-9634d59f63bd4824af9f6766d0de3cfb2023-06-13T20:06:37ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1558-02102022-01-013091592410.1109/TNSRE.2022.31635039745094Epileptic Seizure Detection by Cascading Isolation Forest-Based Anomaly Screening and EasyEnsembleYao Guo0https://orcid.org/0000-0002-9300-5817Xinyu Jiang1https://orcid.org/0000-0002-8518-1415Linkai Tao2Long Meng3https://orcid.org/0000-0001-8140-088XChenyun Dai4https://orcid.org/0000-0002-3056-4339Xi Long5https://orcid.org/0000-0001-9505-1270Feng Wan6https://orcid.org/0000-0002-9359-0737Yuan Zhang7https://orcid.org/0000-0003-2726-2855Johannes van Dijk8Ronald M. Aarts9https://orcid.org/0000-0003-3194-0700Wei Chen10https://orcid.org/0000-0003-3720-718XChen Chen11https://orcid.org/0000-0001-7587-3314Center for Intelligent Medical Electronics, School of Information Science and Technology, Fudan University, Shanghai, ChinaCenter for Intelligent Medical Electronics, School of Information Science and Technology, Fudan University, Shanghai, ChinaDepartment of Industrial Design, Eindhoven University of Technology, Eindhoven, AZ, The NetherlandsCenter for Intelligent Medical Electronics, School of Information Science and Technology, Fudan University, Shanghai, ChinaCenter for Intelligent Medical Electronics, School of Information Science and Technology, Fudan University, Shanghai, ChinaDepartment of Electrical Engineering, Eindhoven University of Technology, Eindhoven, AZ, The NetherlandsCentre for Cognitive and Brain Sciences and the Department of Electrical and Computer Engineering, Faculty of Science and Technology, Institute of Collaborative Innovation, University of Macau, Macau, ChinaCollege of Electronic and Information Engineering, Southwest University, Chongqing, ChinaDepartment of Electrical Engineering, Eindhoven University of Technology, Eindhoven, AZ, The NetherlandsDepartment of Electrical Engineering, Eindhoven University of Technology, Eindhoven, AZ, The NetherlandsCenter for Intelligent Medical Electronics, School of Information Science and Technology, and the Human Phenome Institute, Fudan University, Shanghai, ChinaHuman Phenome Institute, Fudan University, Shanghai, ChinaThe electroencephalogram (EEG), for measuring the electrophysiological activity of the brain, has been widely applied in automatic detection of epilepsy seizures. Various EEG-based seizure detection algorithms have already yielded high sensitivity, but training those algorithms requires a large amount of labelled data. Data labelling is often done with a lot of human efforts, which is very time-consuming. In this study, we propose a hybrid system integrating an unsupervised learning (UL) module and a supervised learning (SL) module, where the UL module can significantly reduce the workload of data labelling. For preliminary seizure screening, UL synthesizes amplitude-integrated EEG (aEEG) extraction, isolation forest-based anomaly detection, adaptive segmentation, and silhouette coefficient-based anomaly detection evaluation. The UL module serves to quickly locate the determinate subjects (seizure segments and seizure-free segments) and the indeterminate subjects (potential seizure candidates). Afterwards, more robust seizure detection for the indeterminate subjects is performed by the SL using an EasyEnsemble algorithm. EasyEnsemble, as a class-imbalance learning method, can potentially decrease the generalization error of the seizure-free segments. The proposed method can significantly reduce the workload of data labelling while guaranteeing satisfactory performance. The proposed seizure detection system is evaluated using the Children’s Hospital Boston-Massachusetts Institute of Technology (CHB-MIT) scalp EEG dataset, and it achieves a mean accuracy of 92.62%, a mean sensitivity of 95.55%, and a mean specificity of 92.57%. To the best of our knowledge, this is the first epilepsy seizure detection study employing the integration of both the UL and the SL modules, achieving a competitive performance superior or similar to that of the state-of-the-art methods.https://ieeexplore.ieee.org/document/9745094/Seizure detection systemunsupervised learningsupervised learningEEGaEEGanomaly detection |
spellingShingle | Yao Guo Xinyu Jiang Linkai Tao Long Meng Chenyun Dai Xi Long Feng Wan Yuan Zhang Johannes van Dijk Ronald M. Aarts Wei Chen Chen Chen Epileptic Seizure Detection by Cascading Isolation Forest-Based Anomaly Screening and EasyEnsemble IEEE Transactions on Neural Systems and Rehabilitation Engineering Seizure detection system unsupervised learning supervised learning EEG aEEG anomaly detection |
title | Epileptic Seizure Detection by Cascading Isolation Forest-Based Anomaly Screening and EasyEnsemble |
title_full | Epileptic Seizure Detection by Cascading Isolation Forest-Based Anomaly Screening and EasyEnsemble |
title_fullStr | Epileptic Seizure Detection by Cascading Isolation Forest-Based Anomaly Screening and EasyEnsemble |
title_full_unstemmed | Epileptic Seizure Detection by Cascading Isolation Forest-Based Anomaly Screening and EasyEnsemble |
title_short | Epileptic Seizure Detection by Cascading Isolation Forest-Based Anomaly Screening and EasyEnsemble |
title_sort | epileptic seizure detection by cascading isolation forest based anomaly screening and easyensemble |
topic | Seizure detection system unsupervised learning supervised learning EEG aEEG anomaly detection |
url | https://ieeexplore.ieee.org/document/9745094/ |
work_keys_str_mv | AT yaoguo epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT xinyujiang epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT linkaitao epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT longmeng epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT chenyundai epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT xilong epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT fengwan epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT yuanzhang epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT johannesvandijk epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT ronaldmaarts epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT weichen epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble AT chenchen epilepticseizuredetectionbycascadingisolationforestbasedanomalyscreeningandeasyensemble |