sidebar
photo
James P. O'Shea
Graduate Student

University of California, Berkeley
Vision Science Program
Visualization Lab
Banks Lab
505 Minor Hall
University of California, Berkeley
Berkeley, CA 94720-2020

joshea (at) cs.berkeley.edu

sidebar
A surgical planning method for functional MRI assessment of language dominance: Influences from threshold, region-of-interest, and stimulus mode

Ralph Suarez, Stephen Whalen, James P. O’Shea, Alexandra J. Golby

Presurgical determination of language laterality is an important step for assessing potential risk of dysfunction resulting from brain resection within or near suspected language areas. Image-based functional MRI (fMRI) methods seek to address limitations to the clinical gold-standard technique by offering a safer, less costly, and non-invasive alternative. In this study we outline a set of protocols for objective determination of langue-specific asymmetry from fMRI activation maps. We studied 13 healthy, right-handed volunteers using a vocalized antonym-generation task. Initially, using the standard threshold-dependent laterality index (LI) procedure, we demonstrated an undesirably high degree of intra-subject variability and indeterminacy in LI value. We addressed this issue by implementing a novel threshold-independent method, resulting in a single, unambiguous LI for each subject. These LIs were then averaged across the group and used to compare functional laterality within the whole hemispheric volumes and six intrahemispheric regions-of-interest (ROIs). We noted that as a result of increased bilateral activation from vocalizations, laterality assessment calculated from the whole hemisphere resulted in insignificant asymmetry. However, by focusing the LI exclusively on the inferior frontal (IFG) and supramarginal gyri (SMG), robust leftward asymmetries were observed. We also examined the influence of stimulus mode on the group mean ROI LI, and observed an increase in IFG asymmetry using visual mode, and in SMG using the auditory mode. Based on these findings, we make recommendations for optimized presurgical protocols.

fig

Group-level fMRI activation maps from all subjects (n=13) performing vocalized antonym-generation in visual mode (left panel) and in auditory mode (right panel). Consistent activation, independent of stimulus mode, was observed bilaterally in the precentral gyrus (PCG), and favoring the left hemisphere in the inferior frontal gyrus (IFG) and the posterior portion of the superior temporal gyrus (TPG). Visual mode activation was observed bilaterally in the middle occipital gyrus (MOG); auditory mode was associated with bilateral activation in the middle portion of the superior temporal gyrus (TTG). Image threshold at P<0.000001, uncorrected.


Research Paper: PDF

A surgical planning method for functional MRI assessment of language dominance: Influences from threshold, region-of-interest, and stimulus mode
Ralph Suarez, Stephen Whalen, James P. O’Shea, Alexandra J. Golby
Brain Imaging and Behavior
June 2008, vol 2, no 2. 59-73.