Abstracts

Rationale: Cortical dysplasias often lead to seizure activity and subsequently epilepsy. Dysplastic tissue is characterized by thickened cortex, and blurred gray matter/white matter (GM/WM) delineation due to disruptions in developmental cortical migration. Subtle dysplasias often go unobserved using imaging techniques. Computational methods of quantifying cortical thickness have been used to characterize gray matter thinning signatures of diseases in many population studies. These measures begin at the WM/GM boundary and measure thickness outward to the GM boundary. Image intensity gradients can also be measured by sampling the image across the WM/GM boundary, assessing the sharpness of the tissue transition. We hypothesized that the addition of these two quantitative techniques can lead to better dysplasia detection. Methods: Two patients with medically retractable epilepsy underwent MR imaging in a work up towards epilepsy surgery. Dysplastic tissue was not observed. Both patients then underwent invasive monitoring with subdural electrodes and follow up resection. High resolution T1 weighted MR images gathered prior to surgery were processed using freesurfer software to measure cortical thickness, and WM/GM boundary intensity gradients. Intensity gradients were calculated by sampling the T1 weighted image in 0.5mm increments beginning 1mm within the GM/WM boundary to 3mm into GM. The slope was then calculated at each point and projected onto the brain s surface (Figure 1 bottom row). Patient cortical thickness data was then compared to ten control subjects. Special attention was given to regions beneath electrodes exhibiting seizure activity. Electrode location maps (figure 1 top row) were created using post-electrode gathered CT and MRI imaging, processed as previously described(1). Results: Both patients exhibited regions near electrodes later measuring epileptic activity that had greater cortical thickness then controls, and less sharply defined GM/WM boundary than other cortex. Following invasive monitoring, seizure-inducing cortex was removed resulting in seizure freedom. Conclusions: The results of this preliminary technique study indicate that subtly defined seizure inducing dysplastic tissue can be detected using computational methods measuring cortical thickness and intensity gradients at the GM/WM boundary. We show two cases where seizure-inducing dysplastic cortex was measurable prior to invasive monitoring. These techniques show promise in epileptic surgical planning. 1.LaViolette PS, Rand SD, Raghavan M, Ellingson BM, Schmainda KM, Mueller WM. 3D Visualization of Subdural Electrodes for Presurgical Planning. Neurosurgery. 2010 2010(In Press).