Revealing brain network hierarchy and molecular correlates in temporal lobe epilepsy: A multimodal neuroimaging study

To investigate hierarchical brain network abnormalities in temporal lobe epilepsy (TLE) using multimodal neuroimaging and explore their molecular mechanisms and cognitive correlates. This study included 52 TLE patients and 49 healthy controls (HCs). Multimodal MRI data were acquired and analyzed for connectivity gradients, Structural Decoupling Index (SDI), and rich-club topology. Gene expression data from the Allen Human Brain Atlas were used to identify molecular correlates. Cognitive function was assessed using the Attention Network Test (ANT). Querol Pascual (2007) ( Querol Pascual, 2007 ) TLE patients showed widespread functional gradient reorganization; while structural gradients remained stable. González Otárula and Schuele (2020) ( González Otárula and Schuele, 2020 ) The Default mode network (DMN) exhibits abnormalities in all imaging modalities. You et al. (2022) ( You et al., 2022 ) Rich-club analysis revealed functional core-periphery dissociation versus structural convergence. Park et al. (2022) ( Park et al., 2022 ) All network metrics showed enrichment for carboxylic acid metabolic processes. Larivière et al. (2020) ( Larivière et al., 2020 ) The SDI robustly predicted all three ANT efficiency measures. This study provides novel evidence for systematic hierarchical network abnormalities in TLE patients through integrated multimodal analysis. Gene enrichment analysis revealed that alterations in structural, functional, and connectivity network metrics were specifically enriched in carboxylic acid metabolic processes, establishing a molecular framework for interpreting TLE network alterations. • Functional hierarchy is disrupted in TLE; DMN shows cross-modal abnormalities. • Functional gradients show core–periphery separation; structural gradients converge. • Imaging-linked genes enrich carboxylic acid metabolism pathways. • SDI and multimodal coupling robustly predict ANT attention performance.