Back

Connectome-wide mega-analysis identifies a reproducible functional network signature of temporal lobe epilepsy

Xie, K.; Chen, J.; Sahlas, E.; Fadaie, F.; Arafat, T.; Rodriguez-Cruces, R.; Naish, M.; Ngo, A.; Dascal, A.; Barnett, A.; Audrain, S.; Pana, R.; Lariviere, S.; Hadjinicolaou, A.; Weil, A. G.; Obaid, S.; Dudley, R.; Schrader, D. V.; Zhang, Z.; Concha, L.; Bernasconi, A.; Bernasconi, N.; Bernhardt, B. C.

2026-03-13 neuroscience
10.64898/2026.03.11.711198 bioRxiv
Show abstract

Resting-state functional magnetic resonance imaging (MRI) studies have reported abnormal intrinsic functional connectivity (FC) across distributed circuits in patients with temporal lobe epilepsy (TLE), indicating a system-level impact of the disorder. However, findings remain inconsistent due to limited sample sizes and methodological heterogeneity, leaving the structural determinants and clinical relevance of FC alterations unresolved. To identify a robust and reproducible FC signature, we conducted a data-driven, connectome-wide mega-analysis in a large multicentre cohort of 652 participants (297 TLE, 73 disease controls, and 282 healthy controls) with multimodal 3T MRI and deep clinical phenotyping. We identified convergent FC reconfigurations at both group and individual levels that preferentially involved densely connected hubs, manifesting as hyperconnectivity in frontoparietal association systems and hypoconnectivity in temporal and paralimbic systems. Integrating structural cortical wiring features further revealed that these extensive alterations were constrained by corticocortical proximity, microstructural similarity, and white matter connectivity. Clinically, the FC phenotype tracked symptom burden and disease progression, informed postsurgical seizure outcome, and distinguished TLE from other focal epilepsies. Collectively, these findings systematically delineate a neurobiologically grounded, hub-centric pattern of intrinsic network disruption in TLE, anchored in temporolimbic and adjacent transmodal systems, with potential utility for individualized phenotypic stratification and outcome prognostication.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.