Spike and wave discharges detection in genetic absence epilepsy rat from Strasbourg and patients with genetic generalized epilepsy.
Absence seizure
Detection
Genetic Absence Epilepsy Rats from Strasbourg
Genetic Generalized Epilepsy
Graph neural network
Spike-wave discharges
Journal
Epilepsy research
ISSN: 1872-6844
Titre abrégé: Epilepsy Res
Pays: Netherlands
ID NLM: 8703089
Informations de publication
Date de publication:
08 2023
08 2023
Historique:
received:
19
10
2022
revised:
02
05
2023
accepted:
13
06
2023
medline:
12
7
2023
pubmed:
26
6
2023
entrez:
26
6
2023
Statut:
ppublish
Résumé
Generalised spike and wave discharges (SWDs) are pathognomonic EEG signatures for diagnosing absence seizures in patients with Genetic Generalized Epilepsy (GGE). The Genetic Absence Epilepsy Rats from Strasbourg (GAERS) is one of the best-validated animal models of GGE with absence seizures. We developed an SWDs detector for both GAERS rodents and GGE patients with absence seizures using a neural network method. We included 192 24-hour EEG sessions recorded from 18 GAERS rats, and 24-hour scalp-EEG data collected from 11 GGE patients. The SWDs detection performance on GAERS showed a sensitivity of 98.01% and a false positive (FP) rate of 0.96/hour. The performance on GGE patients showed 100% sensitivity in five patients, while the remaining patients obtained over 98.9% sensitivity. Moderate FP rates were seen in our patients with 2.21/hour average FP. The detector trained within our patient cohort was validated in an independent dataset, TUH EEG Seizure Corpus (TUSZ), that showed 100% sensitivity in 11 of 12 patients and 0.56/hour averaged FP. We developed a robust SWDs detector that showed high sensitivity and specificity for both GAERS rats and GGE patients. This detector can assist researchers and neurologists with the time-efficient and accurate quantification of SWDs.
Identifiants
pubmed: 37364342
pii: S0920-1211(23)00106-7
doi: 10.1016/j.eplepsyres.2023.107181
pii:
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
107181Informations de copyright
Crown Copyright © 2023. Published by Elsevier B.V. All rights reserved.