Abstracts

RATIONALE:_ Functional MRI (fMRI) has been used to map eloquent areas such as language and motor cortex for use in the presurgical planning of resective surgery for structural lesions. However, for surgeons there is a degree of uncertainy associated with the accuracy of fMRI and rigorous validation studies are required if the surgeon is to rely on fMRI to base surgical decisions. Electrical cortical stimulation is the gold standard for the preoperative mapping of eloquent cortical areas; intraoperative validation using this technique will indicate the accuracy of the fMRI. The aim of this study is to correlate the anatomical site and extent of the fMRI language and motor activation areas by co-registration with the eloquent sites demonstrated with electrical cortical stimulation. METHODS: Ten patients were studied presurgically. Five were tested with fMRI to locate verbal fluency areas and five to locate primary motor regions. The fMRI studies were performed on a 1.5T Seimens machine. For localization of language areas, an original language mapping fMRI paradigm was developed that involved four separate tasks: word generation, category generation, object naming and sentence completion. Motor activation was stimulated with finger movements of each side and then both simultaneously; this was repeated for toe movements. The data was analysed with the aid of the Analysis of Functional Neuroimages? (AFNI) software package. The fMRI activation sites were co-registered with with the StealthStation Image Guided Systems? using a surface matching technique, and then anatomically correlated with the results of Ojemann cortical stimulation? obtained during surgery and/or preoperative subdural grid stimulation. RESULTS: Consistent areas of activation on the fMRI studies were identified in most patients. A close correlation was found between the site of the fMRI and electrical stimulation, but in most cases the latter were more extensive. CONCLUSIONS: The sites of primary motor and language activation identified with our fMRI analyses anatomically correlate well with those of cortical stimulation studies, but are more restricted in area.