Abstracts

RATIONALE: Integrity of visual function is an important consideration in planning resective occipital lobe epilepsy surgery. Resective surgery carries little risk in patients with an established dense hemianopia. The majority of patients being considered for surgery, however, have either normal or minimally affected visual perimetry. The presurgical evaluation of these patients is challenging and often requires intraoperative mapping of the calcarine cortex.
The purpose of this investigation was to develop functional MRI (fMRI) techniques to characterize integrity of functional occipital cortex to assist epilepsy surgical planning. Participants will be able to discuss relationship.
METHODS: 6 control & 21 pts were assessed with 1.5T magnet high speed gradients for echo planar imaging acquired coronally, 20 slice locations identical to the co-registered coronal spin-echo volumes. Visual stimulus fMRI EPI scan duration = 3 minutes 8 seconds (94 data time points). Stimuli presented via mirror from a rear projected 8 Hz reversing checkerboard pattern presented at 20 sec alternating 20 sec grey screen fixation.
Laterality index (LI) calculated from fMRI maps thresholded at t[gt]5. LI = (activated L occipital hemisphere voxels - activated R occipital hemisphere voxels) / (total activated L occipital hemisphere voxels + activated R occipital hemisphere voxels).
Visual perimetery was obtained in 10 patients and related to fMRI activation.
Two LIs were calculated 1). Narrow region surrounding the striate cortex and 2), broader posterior occipito-parietal region.
RESULTS: First, LI correlated with visual perimetery in all 10 pts. Of the 5 pts with normal perimetery, fMRI LI ranged from -0.035 to -0.064, indicating a nearly symmetric fMRI response. A patient with left field deficit had LI of +0.222, indicating a right hemisphere dominant response. Four patients with right field deficits had LI ranging from -0.302 to -0.474, indicating a left lateralized functional MR response.
Second, controls showed a nearly symmetric fMRI response to stimulation (LI -0.07+0.02).
Third, the mean direction of the broad but not narrow LI was significantly related to side of lesion. 15 had lesions in occipital lobe (striate cortex 9, non striate 6) or in the optic radiation (6). The broader LI (occipital-parietal) was more sensitive compared to the narrow (striate cortex) activation and the broader LI was examined. 11 had Rt lesions (LI -0.32+0.09) & 10 had Lt lesions (LI +0.16+0.12) p[lt].0001.
Finally, examining the ability of the LI to classify individual patients, LI correctly predicted contralateral activation in 14 patients. 4 had either a nearly symmetric response (LI range -0.02 to -0.09). 3 of 21 patients had a mislateralization of LI due to diffuseness of the lesion.
CONCLUSIONS: In controls, the fMRI activation task underlying the LI was robust in activating both occipital lobes in a non-lateralized pattern. Furthermore, LI correlates with visual perimetry and was found to be sensitive to the lateralization of the occipital lobe lesion. While robust at a group level, application to individual patients should be cautiously interpreted especially in patients with diffuse lesions. Qualitative assessment of fMRI response patterns indicated dispersion of activation away from the lesion in many patients raises the interesting possibility of reorganization of visual function in occipital cortex.