Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

M2D2 : Maximum-Mean-Discrepancy Decoder for Temporal Localization of Epileptic Brain Activities

Amirshahi, Alireza ; Thomas, Anthony ; Aminifar, Amir LU orcid ; Rosing, Tajana and Atienza, David (2023) In IEEE Journal of Biomedical and Health Informatics 27(1). p.202-214
Abstract

Recent years have seen growing interest in leveraging deep learning models for monitoring epilepsy patients based on electroencephalographic (EEG) signals. However, these approaches often exhibit poor generalization when applied outside of the setting in which training data was collected. Furthermore, manual labeling of EEG signals is a time-consuming process requiring expert analysis, making fine-tuning patient-specific models to new settings a costly proposition. In this work, we propose the Maximum-Mean-Discrepancy Decoder (M2D2) for automatic temporal localization and labeling of seizures in long EEG recordings to assist medical experts. We show that M2D2 achieves 76.0% and 70.4% of... (More)

Recent years have seen growing interest in leveraging deep learning models for monitoring epilepsy patients based on electroencephalographic (EEG) signals. However, these approaches often exhibit poor generalization when applied outside of the setting in which training data was collected. Furthermore, manual labeling of EEG signals is a time-consuming process requiring expert analysis, making fine-tuning patient-specific models to new settings a costly proposition. In this work, we propose the Maximum-Mean-Discrepancy Decoder (M2D2) for automatic temporal localization and labeling of seizures in long EEG recordings to assist medical experts. We show that M2D2 achieves 76.0% and 70.4% of F1-score for temporal localization when evaluated on EEG data gathered in a different clinical setting than the training data. The results demonstrate that M2D2 yields substantially higher generalization performance than other state-of-the-art deep learning-based approaches.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Brain modeling, Decoding, Deep learning, Electroencephalography, Epileptic Seizure, Kernel, Location awareness, Maximum Mean Discrepancy, Non-invasive EEG, Recording, Temporal Localization
in
IEEE Journal of Biomedical and Health Informatics
volume
27
issue
1
pages
13 pages
publisher
IEEE - Institute of Electrical and Electronics Engineers Inc.
external identifiers
  • scopus:85139421136
  • pmid:36136930
ISSN
2168-2194
DOI
10.1109/JBHI.2022.3208780
language
English
LU publication?
yes
id
f552254c-f979-4b1e-a45e-1c153e75ba5d
date added to LUP
2022-12-14 12:54:47
date last changed
2024-04-14 10:02:33
@article{f552254c-f979-4b1e-a45e-1c153e75ba5d,
  abstract     = {{<p>Recent years have seen growing interest in leveraging deep learning models for monitoring epilepsy patients based on electroencephalographic&amp;#x00A0;(EEG) signals. However, these approaches often exhibit poor generalization when applied outside of the setting in which training data was collected. Furthermore, manual labeling of EEG signals is a time-consuming process requiring expert analysis, making fine-tuning patient-specific models to new settings a costly proposition. In this work, we propose the Maximum-Mean-Discrepancy Decoder&amp;#x00A0;(M2D2) for automatic temporal localization and labeling of seizures in long EEG recordings to assist medical experts. We show that M2D2 achieves 76.0&amp;#x0025; and 70.4&amp;#x0025; of F1-score for temporal localization when evaluated on EEG data gathered in a different clinical setting than the training data. The results demonstrate that M2D2 yields substantially higher generalization performance than other state-of-the-art deep learning-based approaches.</p>}},
  author       = {{Amirshahi, Alireza and Thomas, Anthony and Aminifar, Amir and Rosing, Tajana and Atienza, David}},
  issn         = {{2168-2194}},
  keywords     = {{Brain modeling; Decoding; Deep learning; Electroencephalography; Epileptic Seizure; Kernel; Location awareness; Maximum Mean Discrepancy; Non-invasive EEG; Recording; Temporal Localization}},
  language     = {{eng}},
  number       = {{1}},
  pages        = {{202--214}},
  publisher    = {{IEEE - Institute of Electrical and Electronics Engineers Inc.}},
  series       = {{IEEE Journal of Biomedical and Health Informatics}},
  title        = {{M2D2 : Maximum-Mean-Discrepancy Decoder for Temporal Localization of Epileptic Brain Activities}},
  url          = {{http://dx.doi.org/10.1109/JBHI.2022.3208780}},
  doi          = {{10.1109/JBHI.2022.3208780}},
  volume       = {{27}},
  year         = {{2023}},
}