Abstracts

Rationale: The epileptic activity underlying idiopathic generalized epilepsy (IGE) is believed to be related to abnormal thalamo-cortical interactions. Our purpose was to map the organization of the thalamo-cortical network in IGE in vivo using morphometric correlation analysis on MRI data. Methods: We studied 23 patients with IGE and 46 healthy controls. We measured cortical thickness using an unbiased automatic surface-based deformation algorithm (Kim et al., NeuroImage 2005; 27(1):210-21) and calculated the volumes of thalami automatically by applying a non-linear matching algorithm to a probabilistic anatomical atlas (Collins et al., In: Kuba: Image Processing in Medical Imaging 1999; Heidelberg: Springer Verlag; p. 210-23). In each group, we correlated thalamic volume with cortical thickness at each vertex. Significant correlations were interpreted as connections. We assessed differences in thalamo-cortical connectivity between groups by testing differences in correlation between groups at each vertex using linear models. Results: In controls (Figure, upper panel), thalamo-cortical correlations were similar in both hemispheres and encompassed limbic (posterior cingulate, mesiotemporal, fronto-basal, insular, temporopolar), occipito-temporal, as well as fronto-central regions. In IGE patients (Figure, lower panel), the pattern of correlations was similar to controls. Nevertheless, we found increased correlations in frontal (precentral, central), parietal (post-central), temporal (occipito-temporal, lateral temporal) and occipital regions. Decreases in correlations were found in posterior cingulate and mesiotemporal regions bilaterally. Conclusions: Morphometric correlation analysis suggests that thalamo-cortical networks are remodeled in IGE. Increased correlations in fronto-central regions may stem from correlated tissue loss in distant areas due to shared vulnerability to a diffuse insult, or may reflect seizure-related damage. Within the thalamus, pathological processes in somato-motor nuclei may induce secondary damage in limbic nuclei. Disrupted cross talk between thalamic nuclei may in turn induce altered network behavior within the limbic system.