Abstracts

Rationale: Resting-state functional MRI (rs-fMRI) may be used to assess global and regional functional connectivity (FC) of morphometrically and functionally altered brain regions in idiopathic generalized epilepsy (IGE). As regional morphometric abnormalities overlap brain regions functionally participating in the epileptic network, there is reason to believe that the structural abnormalities may give rise to generalized ictal and interictal epileptic discharges by rapid synchronization of functionally connected regions. Hence, in a nonhuman primate model of IGE, we expect rs-fMRI to reveal increased FC between cortical regions and subcortical structures involved in the epileptic network, while connectivity would be decreased in regions or functional networks outside of the epileptic network.Methods: A total of six MRI sessions were performed on three control (CTL) and three epileptic (EPI) baboons (Papio hamadryas anubis, 15 3 kg). The baboons were sedated using intravenous ketamine (5-6 mg/kg/hr) and paralyzed with vecuronium (0.1-0.3 mg/kg). MRIs were performed on a 3T Siemens TIM Trio (Siemens, Erlangen, Germany) using a body radiofrequency coil for transmission and a Siemens 12-channel human head coil for reception. Anatomical MRI was acquired using the MPRAGE pulse sequence with repetition time = 2,100 ms, echo time = 3.1 ms, flip angle = 12 , inversion time = 1,100 ms, field of view = 16x19.2x19.2 cm, 1mm isotropic spatial resolution with two averages. Resting state fMRI (rs-fMRI) acquisitions were performed using a whole-brain blood oxygen level-dependent (BOLD) fMRI sequence with the following parameters: 36 axial slices; an in-plane resolution 1 x 1 mm; slice thickness of 2 mm with no slice gap; TR = 3000 ms; TE = 30 ms; 600 volumes covering the entire brain. Each rs-fMRI scan lasted 30 min. Data processing and analysis were performed using tools from FSL and custom software written in Matlab (MathWorks). We used a voxel-wise approach to map resting-state functional connectivity across the cortex.Results: We found substantial differences in global and regional FC between EPI and CTL baboons (Fig.1). In CTL baboons, the map reflects a distribution resembling the default mode network, with the greatest co-activation in the parieto-temporal association cortices, precuneus and medial frontal cortices. FC is completely disrupted in the EPI baboons, with the greatest connectivity in the orbitofrontal, anterior cingulate, insulae and medial frontal regions, all regions implicated before and during human absence seizures on EEG-fMRI. We also evaluated regional connectivity between areas that were activated on functional imaging scans at rest or during photic stimulation and identified significant correlations between the precuneus or occipital lobes and the SMA in EPI baboons, indicating altered connectivity between these regions.Conclusions: Since rs-fMRIs may be used to assess the global and regional FC differences between EPI and CTL groups, it can provide a new platform for the development and testing of AED and neurostimulation therapies.