Normative functional network localization of gray matter alterations in idiopathic generalized epilepsy

Idiopathic generalized epilepsy (IGE) has traditionally been considered a non-lesional epilepsy syndrome on conventional MRI, yet voxel-based morphometry (VBM) studies have reported subtle gray matter (GM) abnormalities across diverse brain regions. These findings remain spatially heterogeneous and difficult to reconcile. We aimed to characterize the normative functional connectivity patterns associated with reported GM alterations in IGE using functional connectivity network mapping (FCNM), and to explore whether these spatially heterogeneous findings show a reproducible pattern of observed network overlap.

We systematically identified 18 whole-brain VBM studies of IGE, including 111 reported peak coordinates from 1,492 participants. Coordinates reflecting both increased and decreased GM were pooled in the primary analysis to characterize normative connectivity patterns associated with reported GM alterations irrespective of alteration direction. Using resting-state functional MRI (rs-fMRI) data from a normative discovery cohort (Human Connectome Project, N = 1,093), we mapped the functional connectivity of these GM coordinates to identify a convergent IGE-related network. The reproducibility of the observed spatial pattern was explored using an independent, cross-scanner validation cohort (Southwest University Adult Lifespan Dataset, N = 329) and two alternative seed sizes (1-mm and 7-mm radii).

Reported GM alterations in IGE were associated with an observed distributed normative functional connectivity pattern involving the bilateral middle temporal gyri, bilateral thalami, left superior frontal gyrus, and right putamen. In the HCP discovery analysis, this network showed the greatest observed overlap with the subcortical network (overlap proportion = 42.77%) and secondary correspondence with the default mode network (DMN; overlap proportion = 11.06%). Analysis using the independent SALD connectome showed a similar descriptive pattern, with the greatest observed overlap again involving the subcortical network (overlap proportion = 43.98%), whereas DMN correspondence was weaker.

Reported GM alterations in IGE converge most robustly onto a subcortical normative functional network involving the thalamus and putamen, with additional DMN involvement observed primarily in the HCP discovery analysis. These findings provide preliminary, descriptive evidence that reported GM alterations in IGE are associated with a reproducible subcortical normative connectivity pattern across two connectome datasets, although the available analyses do not establish above-chance or disease-specific network convergence.