Abstracts

RATIONALE: Epilepsy due to FCD is usually refractory to AED, but can be controlled if the lesion is resected surgically. FCD may be subtle and extend beyond the limits of resection defined by visual analysis of MRI, and the residual malformation is often the reason for poor surgical outcome. Pre-operative neuroimaging evaluation, particularly MPR and CR, plays an important role in the diagnosis of subtle lesions, and potentially it can also help to improve anatomical localization and extension of FCD. The objective of this study was to establish the contribution of multiplanar reconstruction (MPR) and curvilinear reformatting (CR) to the MRI investigation of focal cortical dysplasia (FCD).
METHODS: We selected 12 patients with diagnosis of FCD that were being evaluated for epilepsy surgery. Neuroimaging evaluation was prospectively reviewed to establish the contribution of each post processing techniques for the diagnosis and identification of the anatomical location and extension of the FCD. One of us analyzed systematically, blinded to the clinical data, the plain films using 6mm sagittal T1, 3mm coronal IR and T2, and 3mm axial T1, T2 and FLAIR images, followed by MPR and CR on volumetric T1 1mm isotropic voxel acquisition. We analyzed the results in two stages: firstly, we compared the findings of plain films to MPR; and secondly we compared the findings of CR to plain film plus MPR. For comparison of the extent of FCD between MPR and CR, the lesion shown by MPR was manually delineated and then compared directly to the 3D curvilinear images.
RESULTS: Plain films of high resolution MRI showed the lesion in six (50%) patients whereas MPR and CV analysis in all patients. MPR provided a better discrimination of lesion localization and its relationship to other cerebral structures in five (83%) of the six patients with lesion identified on plain films of MRI. CR improved lesion localization over MPR in one patient. The lesion showed by CR was larger than the one seen on MPR in three patients. In two patients CR showed that areas suspected as abnormal on MPR was most likely volume averaging.
CONCLUSIONS: Both MPR and CR add to the presurgical assessment of patients with FCD by improving lesion diagnosis and localization. In addition, CR establishes more precisely the anatomical distribution of the lesion, allowing the identification of abnormal areas not shown on the plain MRI film or MPR ascertainment.
Support: FAPESP