Meeting Banner
Abstract #3659

Effectiveness of Four Different Clinical Language Paradigms for Language Lateralization: A ROI Analysis

Domenico Zac1, Joshua Nickerson2, Gerard Deib1, Jay J. Pillai1

1Division of Neuroradiology Johns Hopkins University School of Medicine, Baltimore, MD, United States; 2Division of Neuroradiology, The University of Vermont School of Medicine, Burlington, VT, United States

Language hemispheric dominance determined by BOLD fMRI can depend on the paradigm as well as the region of interest (ROI) considered for the lateralization index (LI) computation. In this retrospective study of 41 brain tumor patients referred for presurgical language mapping we compared the effectiveness of lateralization of four language tasks in anatomical and functional language ROIs. Two expressive tasks, rhyming and silent word generation, were the best lateralizing tasks in expressive ROIs, whereas the receptive and semantic tasks, sentence completion and listening comprehension, did not perform better than the expressive tasks for lateralization in receptive ROIs.

Keywords

acquisition activation affecting aims among anatomic anatomical anatomically anatomy approved approximated axial better binder board bold boundaries brain calculation choice clinical clinically cluster clusters coefficient completion comprehension concordance confirms considered consistently correlation cortex defined design designed determination determine determined division dominance dominant drawn dual effectiveness effort eloquent encompass essentially expressive extensive extent feasibility findings frame frontal functional functionally general gold good grade handed hemisphere hemispheric hertz included independent index influence institutional interaction intraclass invasive johns language lateralization lesions listening lobar lobe lobes manually mapping maps margin matrix medicine model need neurology optimal overlay pair pannier paper paradigm paradigms parietal patient patients populations position post posterior preoperative presentation previous primarily prism processed rater raters rather receptive referred remaining replaced reported resection resolution retrospective reveal review risk school score segmenting semantic sentence series significantly slices sodium software spatial statistical studies system tasks temporal thick threshold trend trio tumor tumors validation volume volumetric word