MRI Correlates of Focal Cortical Dysplasia: A Study with Multimodal Image Analyses.
Abstract number :
1.226
Submission category :
Year :
2000
Submission ID :
1384
Source :
www.aesnet.org
Presentation date :
12/2/2000 12:00:00 AM
Published date :
Dec 1, 2000, 06:00 AM
Authors :
Sabine Rona, Imad M Najm, Zhong Ying, Eric L LaPresto, Shiyang Wang, William E Bingaman, Hans O Lnders, The Cleveland Clin Fdn, Cleveland, OH.
RATIONALE: Focal cortical dysplasia (CD) is a frequent cause of medically refractory partial epilepsy. Magnetic resonance imaging (MRI) changes are often subtle and may be missed on conventional 2D images. This study uses semi-automated interactive multimodal image analyses to establish the MRI correlates of CD. METHODS: 14 patients (5M and 9F, mean age 20 years) with histologically proven CD evaluated with high-resolution MRI and subdural grid electrodes who underwent resective surgery at CCF between 8/97 and 9/99 were analysed retrospectively with regard to MRI characteristics, and the results were correlated with histological findings. Pre-operative T1 and FLAIR images and post-operative T1 images were co-registered to allow for spatial matching. Structural and signal abnormalities were assessed with 1) visual analysis, including a threshold-based display method for FLAIR images, and 2) quantitative FLAIR signal analysis using volumes of interest (VOIs) corresponding to the total area of resection as well as to subregions with CD. RESULTS: Frontal lobe resection (FR) was performed in 10 and temporal lobe resection (TR) in 4 patients. Dual pathology (hippocampal sclerosis) was found in 3/4 TR. Balloon cells (BC) were seen in 4/10 patients with FR. 2D visual image analysis showed structural abnormalities in 12 and FLAIR signal changes in 11 patients. 3D surface analysis provided additional findings in 4 patients; FLAIR threshold analysis revealed areas of signal increase outside the frontal lobe in 5/10 patients with FR, in 4/5 cases correlated with pre-operative EEG abnormalities. Focal cortical thickening with FLAIR signal increase was present only in patients with CD associated with BC. Quantitative VOI analysis did not show a significant FLAIR signal increase for the total resected area or subregions with CD without BC, but increases of 14 and 19% were seen in 2 regions with BC. CONCLUSIONS: 1) Focal cortical thickening associated with FLAIR signal increase is highly predictive of the presence of BC in CD. 2) 3D surface and FLAIR threshold analyses are useful adjunctive techniques for the detection of additional morphological abnormalities in patients with CD.