Abstracts

To obtain reliable measurements of cortical thickness in JME patients, we adopted a fully automatic method to reconstruct the pial surface, called CLASP (Constrained Laplacian-based Anatomic Segmentation using Proximity) algorithm, which provides a topology-preserving cortical surface deformation method and compared the cortical thickness between JME patients and age/sex matched normal controls., We selected 19 JME patients (mean age[plusmn]S.D.= 22.6[plusmn]4.9 years, 6 male) who were diagnosed by the International League Against Epilepsy classification of epilepsy and epileptic syndrome and closely age- and sex-matched 19 normal controls (6 male) who had no familial and personal history of neurological or psychiatric diseases, and no head trauma. In each SPGR (Spoiled Gradient Recalled) brain, the inner (i.e. GM/WM boundary) and outer (i.e. GM/CSF boundary) surface was extracted by CLASP. The CLASP algorithm consists of several stages as follows: 1) Acquired T1 MR images are preprocessed by intensity inhomogeneity correction, 2) spatial normalization to stereotaxic space, 3) preprocessed images are classified into GM, WM, and CSF tissues, 4) the WM surface is reconstructed by deforming a spherical polygon model to the white matter boundary, 5) a Laplacian field is generated between the WM surface resampled to voxel space and a skeletonized CSF fraction image, and 6) the GM surface is initiated from the WM surface and is expanded to the boundary between GM and CSF along the Laplacian field. Given a surface representing the GM/CSF boundary and another representing the GM/WM boundary, the cortical thickness measure could be performed. Among various metrics for thickness measure, [italic]T[/italic]-link was used for thickness measure, which means the distance from a vertex on the outer surface to the corresponding vertex on the inner surface, as defined by the linkages used in the two-surface deformation of the CLASP algorithm., There were no structural or signal abnormalities in SPGR, T2-weighted, and Fluid Attenuated Inversion Recovery MRI in both JME patients and normal controls on visual inspection. In JME, cortical thickness was decreased in bilateral medial and superior frontal gyri, left middle frontal gyrus, left superior temporal gyrus, bilateral middle temporal gyri, right inferior temporal gyrus, and right fusiform gyrus compared to normal controls (t-test, p[lt]0.001). There was no cortical area showing increased thickness in the JME brains., The JME patients showed reduced cortical thickness in bilateral frontal and temporal lobes, which may be developmental anomaly or consequences of repetitive seizures.,