Improved reliability of intraoperative language testing through pre-operative baseline linguistic scores and baseline object naming

Abstract

Abstract Visual object naming (vON) is the most commonly applied linguistic test during awake surgeries with electrical stimulation mapping. Little is known about the predictive value of general preoperative linguistic and cognitive function for the intraoperative object naming ability of the patient. We aimed at analyzing these correlations, in order to potentially define cut-off values for when intraoperative vON tasks may no longer be gainful. Also, we aimed to assess the benefits of patient-specific tailoring of object-images. 46 patients with left-sided perisylvian tumors, scheduled for awake surgeries, underwent a preoperative workup including a comprehensive test battery for general linguistic function, a cognitive function test, and an object naming task, employing a set of objects that was validated prior in an inhouse study. For intraoperative use, the initial set of 80 objects was tailored down to a slack of objects, each patient could reproducibly name. Correlations between the respective tests were drawn using multivariate analyses. On average, patients were only able to correctly name 81% of the original validated baseline set of objects (range from 16%-88%). Aachen Aphasia Test (AAT) scores and DemTect test scores (DS) correlated tightly with the vON. Patients with initial AAT scores of < 80%, DS of < 50%, and baseline vON scores of < 50% were no longer able to reliably comply with the intraoperative task. Patient specific tailored sets of objects improve the reliability of intraoperative object naming tasks. An elaborate preoperative language evaluation will give an estimate of the expected object naming ability of the patient and will help put into perspective intraoperative performances. There may be cut-off values in a-priori language function that may disqualify for awake surgeries.