Brain network construction and analysis for epilepsy: A methodology review.

Abstract

Epilepsy is increasingly understood as a network disorder, prompting widespread application of graph-theoretic methods to electrophysiological and neuroimaging data. The resulting methodological landscape-spanning functional, structural, and effective connectivity, together with dynamic and multimodal extensions-has grown fragmented. Following PRISMA guidelines, we screen Web of Science, IEEE Xplore, PubMed, and Scopus, and reorganize the literature around five prevalent research tasks: seizure prediction, seizure detection, seizure type classification, clinical correlation analysis, and epileptic foci localization. This task-oriented structure differs from conventional modality-centric reviews. Section 3 consolidates the conceptual foundations and construction methods for functional, structural, and effective networks, together with their dynamic and multimodal extensions, and clarifies that graph theory metrics must be interpreted according to network type. Sections 4 and 5 survey methodological adoption across the five tasks, characterize the cross-task distribution of network strategies, and analyze the convergence of dynamic modeling, graph neural networks, and multimodal fusion, as well as current gaps in clinical practice.