Abstracts

Rationale: fMRI activation of the hippocampi may yield information for surgical planning. As memory fMRI is methodologically challenging, other indirect methods of studying hippocampal activation (HA) may yield useful data. We compared HA and hippocampal LI during a language and a verbal memory task in a pediatric epilepsy population. Methods: We examined HA during two separate EPI BOLD 3T block design tasks: a language task and a verbal memory task. The language task was an age-level adjusted auditory description decision task. The verbal memory task consisted of task of paired association learning with word pair stimuli. Six blocks were presented with each block consisting of three trials [learning, distraction (baseline), and recall]. Five patients (mean age: 13.6, range: 10.5-16.7 years) completed both paradigms. Imaging processing and statistical analyses were conducted in SPM8. Image normalization and segmentation was done using subjects high-resolution T1 images with the VBM8 Toolbox in SPM8, then applying those parameters to the EPI images. A field map was also acquired and used to correct distortion. The region of interest (ROI) for bilateral hippocampi was based on the Anatomical Atlas Library in the Wake Forest PickAtlas. We calculated a laterality index (LI) for the hippocampal ROI using the LI Toolbox. Results: On an individual basis, 60% (3 of 5) of the patients demonstrated HA during the language task (all analyses thresholded at p=0.05, uncorrected). Eighty percent (4 out of 5) demonstrated HA during the verbal memory task (40% for learning trial and 60% for recall trial). A group map (p=0.05, uncorrected) for the verbal memory task rendered a cluster of 13 voxels in the right hippocampus for the learning trial, and 37 voxels in the right and 10 in the left hippocampus for the recall trial. No significant clusters of activation survived for the language task group map. Regarding hippocampal laterality, for the language task 2 patients demonstrated right lateralization and 1 left (for the 2 additional patients LI could not be computed due to limited activation). For the verbal memory recall trial 3 patients had left lateralization, 1 right, and 1 bilateral. Of the 3 patients where hippocampal LI could be calculated for both language and memory tasks, 2 showed the same pattern of LI across all tasks (1 left, 1 right). Conclusions: The preliminary results demonstrate that this verbal learning paradigm reliably activates the hippocampus on an individual basis and does so better than the language task, particularly at the group level. Due to limited HA during the language task, it was difficult to compare LI among tasks; however, when LI could be calculated for both tasks, the LI was consistent among tasks (2 out of 3 cases). We plan to add additional subjects in order to further investigate HA and LI during both tasks. We will also examine effect of pathology and seizure foci as well as age, duration of epilepsy, and other clinical variables, as these factors will likely influence HA.