ASYMMETRIC WHITE MATTER VOLUMES AND MEASUREMENT OF CORTICAL THICKNESS IN FOCAL EPILEPSY
Abstract number :
2.315
Submission category :
Year :
2004
Submission ID :
4764
Source :
www.aesnet.org
Presentation date :
12/2/2004 12:00:00 AM
Published date :
Dec 1, 2004, 06:00 AM
Authors :
1Douglas Rose, 2Venkat Tanneru, 2Lydia Regis, 3Blaise Jones, 1Deborah Holder, and 1Mary Bare
Children with clinically focal seizures from cortical dysplasia (CD) can still have relatively generalized EEG epileptiform discharges (ED). CD can be detected on MRI as altered cortical thickness. Gyral foldings in the plane of MRI sections can appear thick, limiting visual detection of actual altered thickeness. We hypothesize: 1. Segmentation of 3D reconstruction into gray and white matter (WM), and subsequent quantitation, can improve detection. 2. Lateralized differences in WM volumes indicate regions of prior neuronal loss or abnormal development that may be regions where seizures start. 3. Measurement of cortex perpendicular to mantle surface can increase detection of altered cortical thickness. 14 patients 3.5 to 20 y old with intractable clinically focal onset epilepsy had EEG and thin slice SPGR isotropic voxel studies with 1.5T GE scanner. 3D MRI were reconstructed with Curry software. Reconstructions were placed in Tailarach orientation, then segmented into gray and white matter after transition pixel intensity was optimized. WM volumes were calculated for whole brain and for each hemisphere. White matter hemisphere volumes (WMHV) were considered significantly different only when difference was [gt] 10 % of total WM volume. Measurement points for cortical thickness were established on both external (cortex-CSF) and internal (cortex-white matter) surface of the 3D cortical mantle for completeness (40-50K points). Cortical thickness was measured perpendicular to the surface at each point and color-coded. All 3D surfaces were visually reviewed for thinner and thicker regions. Five patients had bilateral ED. Three had no ED recorded. Three had left and 4 had right hemisphere ED. MRI scans were formally read as normal in 4 with bilateral ED and in one each with left hemisphere, right hemisphere, or no recorded ED. The MRIs of the remaining 7 were formally read as focal cortical dysplasia. For patients with bilateral ED, no ED, and/or MRI scans formally read normal, WMHV were more often symmetric or left smaller than right. For patients with right ED, WMHV were symmetric. At certain regions (major fissures, operculum, and at tentortium), the proximity of two contiguous folded gyral surfaces was measured as correspondingly thicker. Three of 7 abnormal MRIs were identified; 3 of normal MRIs had focal regions specified as thicker by the algorithm. White matter hemispheric volume asymmetries did not directly reflect side of epileptiform discharges. In this limited population, WMHV were symmetric or more often left smaller than right. Measurements perpendicular to the cortical surface identified some regions of abnormal mantle thickness. Improvements in the algorithm to evaluate two contiguous cortical surfaces are required to improve detection.