Seizure detection using ultra‐long‐term subcutaneous electroencephalography: A deep learning CNN ‐ BiLSTM approach
Epilepsia · 2026-01 · 67 (1) · 437-447
DOI: 10.1111/epi.18652 ↗Kurz gefasst
Fachartikel in der Zeitschrift Epilepsia, online erschienen am 07.10.2025. Verfasst von Sihyeong Park und 10 weiteren Autor:innen. Beteiligte Einrichtung: Department of Neurology, Mayo Clinic, Rochester, Minnesota, USA. Hauptthemen laut MEDLINE: Deep Learning, Electroencephalography, Seizures, Neural Networks, Computer. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 24 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: Epilespy Foundation of America, NINDS NIH HHS, Mayo Clinic Neurology AI Program.
Abstract (Originalsprache)
Objective
This study reports development and validation of a deep learning seizure detection algorithm for two‐channel subscalp electroencephalographic (EEG) recordings. Ultra‐long‐term monitoring of people with epilepsy may produce new insights into the timing and pattern of their seizures and may pave the way for novel therapeutic options. Although EEG is the accepted standard for epilepsy monitoring, ultra‐long‐term EEG recordings generate a massive quantity of data and are not reasonably reviewable in full by human readers.
Methods
The convolutional neural network‐bidirectional long short‐term memory (CNN‐BiLSTM) hybrid algorithm uses nine layers, operating on channel spectrograms 5 min in length with 50% overlap. Retrospective subscalp EEG data from 16 patients recorded at three centers were available for algorithm development and testing. EEG was recorded for a median of 63 days (range = 9–508), and a median of 17 seizures (range = 0–96) were recorded. Training data were augmented with scalp EEG seizures, and performance was benchmarked against a conventional spectral power classifier algorithm. We also evaluated an intrapatient training–testing approach where the algorithm was trained on the first 45% of data.
Results
The CNN‐BiLSTM trained on scalp EEG and subscalp EEG achieved an area under the receiver operating characteristic curve (AUROC) of .98 and an area under the precision–recall curve (AUPRC) of .50, which correspond to 94% sensitivity with 1.11 false detections per day. The same model trained on iEEG achieved only AUROC = .94 and AUPRC = .36. The conventional spectral band power detector achieved AUROC = .93 and AUPRC = .38. The CNN‐BiLSTM detector trained on the earliest half of the subscalp EEG data achieved AUROC = .93 and AUPRC = .37, corresponding to 87% sensitivity and 5.9 false detections per day. Significance High sensitivity and specificity are possible in automated seizure detection in two‐channel subscalp EEG data using a CNN‐BiLSTM framework. Performance of the detector is superior using subcutaneous EEG data for training rather than intracranial EEG, but addition of scalp EEG seizures for training was helpful.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Neurology, Mayo Clinic, Rochester, Minnesota, USA.
- Institute of Psychiatry, Psychology, and Neuroscience, King's College London, London, UK.
- Epilepsy Centre, King's College Hospital NHS Foundation Trust, London, UK.
- Centro de Estudos Egas Moniz, Faculty of Medicine, University of Lisbon, Lisbon, Portugal.
- Department of Physiology and Biomedical Engineering, Mayo Clinic, Rochester, Minnesota, USA.
- UNEEG, Lillerod, Denmark.
ORCID-Profile: Sihyeong Park, Pedro F Viana, Jonas Duun-Henriksen, Nicholas Gregg, Gregory A Worrell, Benjamin H Brinkmann
Förderung
- Epilespy Foundation of America
- NINDS NIH HHS
- Mayo Clinic Neurology AI Program
- National Institute of Neurological Disorders and Stroke
Zitieren
APA
Park, S., Clark, J. S., Viana, P. F., Cui, J., Duun-Henriksen, J., Mandrekar, J., Gregg, N., Kremen, V., Worrell, G. A., Richardson, M. P., & Brinkmann, B. H. (2026). Seizure detection using ultra‐long‐term subcutaneous electroencephalography: A deep learning CNN ‐ BiLSTM approach. Epilepsia, 67(1), 437-447. https://doi.org/10.1111/epi.18652
Vancouver
Park S, Clark JS, Viana PF, Cui J, Duun-Henriksen J, Mandrekar J, et al. Seizure detection using ultra‐long‐term subcutaneous electroencephalography: A deep learning CNN ‐ BiLSTM approach. Epilepsia. 2026;67(1):437-447. doi:10.1111/epi.18652
BibTeX
@article{Park2026,
author = {Park, Sihyeong and Clark, Jordan S and Viana, Pedro F and Cui, Jie and Duun-Henriksen, Jonas and Mandrekar, Jay and Gregg, Nicholas and Kremen, Vaclav and Worrell, Gregory A and Richardson, Mark P and Brinkmann, Benjamin H},
title = {Seizure detection using ultra‐long‐term subcutaneous electroencephalography: A deep learning CNN ‐ BiLSTM approach},
journal = {Epilepsia},
year = {2026},
volume = {67},
number = {1},
pages = {437--447},
doi = {10.1111/epi.18652}
}
EUVIMED