EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Two‐stage spatiotemporal graph neural network framework for seizure onset zone localization

Julie Le Tallec iD, Ralph Saber iD, Amirhossein Jahani iD, Guillaume Theaud iD, Sami Obaid iD, Dang Khoa Nguyen iD, Elie Bou Assi iD

Epilepsia · 07.11.2026

DOI: 10.1002/epi.70573 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Epilepsia, online erschienen am 07.11.2026. Verfasst von Julie Le Tallec und 6 weiteren Autor:innen. Beteiligte Einrichtung: Centre de Recherche du Centre hospitalier de l'Université de Montréal, Montreal, Quebec, Canada. Die Arbeit nennt 42 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Natural Sciences and Engineering Research Council of Canada, Fonds de Recherche du Québec - Santé, Canada Excellence Research Chairs, Government of Canada.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objective

Accurate localization of the seizure onset zone (SOZ) is critical for the presurgical evaluation of drug‐resistant epilepsy. Intracranial electroencephalography captures high‐resolution seizure dynamics, but interpreting complex spatiotemporal patterns across implanted electrodes is clinically challenging. We propose a two‐stage, uncertainty‐aware framework for SOZ localization that integrates patient‐specific seizure modeling with cross‐patient probabilistic ranking to estimate the likelihood of each electrode belonging to the SOZ.

Methods

The framework was evaluated on a bi‐institutional cohort of 36 patients (20 with good surgical outcome, 16 with poor surgical outcome), comprising 184 seizures. In the first stage, a patient‐specific spatiotemporal graph neural network modeled seizure onset as a network‐level transition using directed functional connectivity graphs derived from burst‐integrated electrophysiological features. Trained solely to detect preictal‐to‐ictal transitions, this model generated electrode probability trajectories without SOZ supervision. In the second stage, these trajectories were summarized into interpretable descriptors and used to train a patient‐agnostic classifier assigning SOZ likelihood scores to electrodes. Evaluation and hyperparameter optimization were performed using nested leave‐one‐patient‐out cross‐validation restricted to good outcome patients, ensuring strict patient‐level separation between training, validation, and test data.

Results

Seizure onset detection was stable across patients, with comparable binary cross‐entropy values between outcome groups (median = .495 vs. .438, p = .108). In good outcome patients, SOZ electrodes were prioritized among top‐ranked contacts (area under the curve = .86 ± .06), with all patients exhibiting at least one SOZ electrode within the top positions. In poor outcome patients, predictions exhibited reduced spatial separability, lower likelihood magnitudes, and near‐zero confidence gaps, reflecting attenuated model confidence. Significance This two‐stage framework integrates spatial interactions, temporal dynamics, and probabilistic ranking within a pipeline. By providing graded SOZ likelihood estimates and uncertainty‐aware prioritization, the approach is a decision‐support tool to assist hypothesis refinement during presurgical evaluation rather than as a standalone diagnostic system.

Schlagwörter

Stichwörter der Autor:innen

drug‐resistant epilepsygraph neural networksintracranial EEGpresurgical evaluationseizure onset zone localization

Autor:innen und Institutionen

  • Centre de Recherche du Centre hospitalier de l'Université de Montréal, Montreal, Quebec, Canada.
  • Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland.
  • Department of Neuroscience, Université de Montréal, Montreal, Quebec, Canada.
  • Obaid Neuroimaging for Surgical Epilepsy Treatment Lab, Centre de Recherche du Centre Hospitalier de l'Université de Montréal, Montreal, Quebec, Canada.
  • Division of Neurosurgery, Centre Hospitalier de l'Université de Montréal, Montreal, Quebec, Canada.
  • Department of Surgery, Université de Montréal, Montreal, Quebec, Canada.

ORCID-Profile: Julie Le Tallec, Ralph Saber, Amirhossein Jahani, Guillaume Theaud, Sami Obaid, Dang Khoa Nguyen, Elie Bou Assi

Förderung

  • Natural Sciences and Engineering Research Council of Canada
  • Fonds de Recherche du Québec - Santé
  • Canada Excellence Research Chairs, Government of Canada
  • CRCHUM
  • CIHR
  • Canadian Institutes of Health Research

Zitieren

APA

Le Tallec, J., Saber, R., Jahani, A., Theaud, G., Obaid, S., Nguyen, D. K., & Bou Assi, E. (2026). Two‐stage spatiotemporal graph neural network framework for seizure onset zone localization. Epilepsia. https://doi.org/10.1002/epi.70573

Vancouver

Le Tallec J, Saber R, Jahani A, Theaud G, Obaid S, Nguyen DK, et al. Two‐stage spatiotemporal graph neural network framework for seizure onset zone localization. Epilepsia. 2026. doi:10.1002/epi.70573

BibTeX
@article{LeTallec2026,
  author = {Le Tallec, Julie and Saber, Ralph and Jahani, Amirhossein and Theaud, Guillaume and Obaid, Sami and Nguyen, Dang Khoa and Bou Assi, Elie},
  title = {Two‐stage spatiotemporal graph neural network framework for seizure onset zone localization},
  journal = {Epilepsia},
  year = {2026},
  doi = {10.1002/epi.70573}
}
Kostenlos starten →