Abstracts

Purpose: First, to compare the cortical topography of language-specific cortex as delineated by the invasive [ldquo]gold standard[rdquo] procedure (direct electrocortical stimulation (ECS)) and by a non-invasive functional imaging procedure (Magnetoencephalography (MEG)). Second, to test the contributions of both methods in identifying [ldquo]essential[rdquo] language cortex in patients undergoing brain surgery on the basis of their postsurgical linguistic outcome. In 47 patients undergoing brain surgery, language cortex was delineated using MEG recordings and ECS. The cortical topography drawn after MEG and ECS was compared on a case by case basis to test the degree of spatial overlap. In a subgroup of 14 patients, pre and postsurgical linguistic performance was available. The spatial overlap of the resected area with the language maps derived after MEG and ECS mapping was then estimated. The relative merits of these two techniques in identifying [ldquo]essential[rdquo] language cortex was estimated after reviewing the effects of the resection on the postsurgical linguistic performance. In our 47 patients, 63 language-specific cortical sites were defined by MEG recordings and 55 out of those (87%) were verified using ECS (within [lt] 1cm). Repetition errors were the most common finding (78%), followed by naming errors (37%). After reviewing post-surgical MRIs and linguistic outcomes in a subgroup of 14 patients, resection of MEG and ECS defined language areas had a similar ability to predict linguistic regression after surgery (i. e., naming/fluency worsening). Direct recordings of neurophysiological signals using MEG can help to identify [ldquo]essential[rdquo] language cortex and be a valid clinical tool to prevent functional morbidity in patients undergoing brain surgery. (Supported by Vivian L. Smith Center for Neurologic Research, NINDS grant NS37941 to A.C. Papanicolaou.)