Abstracts

Rationale: Although understanding the lateralization of language functions is important for planning epilepsy surgery, the method and threshold used to determine laterality is debated. This study compares four methods used to determine language dominance.Methods: We studied 50 epilepsy patients (ages 8-55) with a left focus characterized by history, vEEG, and MRI. Functional imaging was conducted using whole brain 3T fMRI (EPI BOLD) with an auditory description decision paradigm employing a box car design. Data was analyzed with SPM2. Whole brain and mask analyses were used. The mask covered inferior frontal gyrus (IFG). We compared four laterality methods. 1) Visual rating was conducted based on the whole brain analysis focusing on IFG at thresholds of p<0.001 uncorrected (t=3.15) and p<0.01 uncorrected; 2) we used the SPM2 LI Toolbox to determine laterality of patients at a threshold of t=3.15 (p=0.001); 3) we used a data driven method for determining laterality by basing the calculation on an adaptive threshold (the tmax 1/2 activated voxels); 4) we created a second data driven method for determining laterality basing calculation on 2x t adaptive threshold (the top tmax ¾ of activated voxels). In each method, a minimum of 5 voxels were required in one region to calculate the LI. LI was then calculated = [L-R]/[L+R] with LI < -0.2, right dominant; LI >