EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Functional connectivity methods for detecting seizure onset zone in focal epilepsy: an approach through temporal graphs and community detection

Pablo A. Osorio Marulanda, Juan D. Martínez-Vargas, Lina M. Sepúlveda-Cano, Alejandro Peña

PeerJ Computer Science · 28.07.2026 · 12 · e3869

DOI: 10.7717/peerj-cs.3869 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PeerJ Computer Science, online erschienen am 28.07.2026. Verfasst von Pablo A. Osorio Marulanda und 3 weiteren Autor:innen. Die Arbeit nennt 56 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Universidad EAFIT.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Focal epilepsy, a prevalent neurological disease with seizures from localized brain regions, poses significant diagnostic and treatment challenges. Precise seizure onset zone (SOZ) localization is crucial for surgery in drug-refractory cases. However, traditional static or time-averaged functional connectivity (FC) models often fail to capture the dynamic brain processes inherent in epileptic activity. This study introduces a novel methodology for SOZ identification using dynamic graph representations of stereo-electroencephalography (SEEG)-derived FC. We constructed frequency-specific (theta, alpha, beta, gamma) temporal networks using Amplitude Envelope Correlation (AEC) and Phase-Locking Value (PLV), incorporating spatial distance normalization. These networks were thresholded, binarized, and analyzed using an ensemble of community detection algorithms. A key innovation is our novel intra-community density score with size regularization (IDS), designed to prioritize communities indicative of epileptogenic activity. Validation on simulated datasets (Epileptor model) yielded an area under the curve (AUC) of 0.74. On clinical SEEG from 19 focal epilepsy patients, our framework demonstrated significant utility, particularly when analyzing the low gamma band with 6-s epochs using the AEC method. Stratified analyses revealed strong performance in specific subgroups, achieving an AUC of 0.76 in patients with mesial temporal lobe lesions, highlighting the method’s potential for targeted clinical applications. While the average AUC across the entire heterogeneous patient cohort was 0.50, these specific findings underscore the importance of patient-specific factors and tailored analytical approaches. Our results underscore the importance of temporal network analysis in capturing ictal dynamics. This study demonstrates that dynamic, graph-based FC analysis, incorporating advanced community detection and our novel IDS metric, offers a promising strategy to enhance presurgical SOZ evaluation in focal epilepsy.

Förderung

  • Universidad EAFIT

Zitieren

APA

Marulanda, P. A. O., Martínez-Vargas, J. D., Sepúlveda-Cano, L. M., & Peña, A. (2026). Functional connectivity methods for detecting seizure onset zone in focal epilepsy: an approach through temporal graphs and community detection. PeerJ Computer Science, 12, e3869. https://doi.org/10.7717/peerj-cs.3869

Vancouver

Marulanda PAO, Martínez-Vargas JD, Sepúlveda-Cano LM, Peña A. Functional connectivity methods for detecting seizure onset zone in focal epilepsy: an approach through temporal graphs and community detection. PeerJ Computer Science. 2026;12:e3869. doi:10.7717/peerj-cs.3869

BibTeX
@article{Marulanda2026,
  author = {Marulanda, Pablo A. Osorio and Martínez-Vargas, Juan D. and Sepúlveda-Cano, Lina M. and Peña, Alejandro},
  title = {Functional connectivity methods for detecting seizure onset zone in focal epilepsy: an approach through temporal graphs and community detection},
  journal = {PeerJ Computer Science},
  year = {2026},
  volume = {12},
  pages = {e3869},
  doi = {10.7717/peerj-cs.3869}
}
Kostenlos starten →