Improving Accuracy of Language Area Mapping by Passive Functional Mapping and Cortico-Cortical Evoked Potentials
Abstract number :
3.184
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2019
Submission ID :
2422082
Source :
www.aesnet.org
Presentation date :
12/9/2019 1:55:12 PM
Published date :
Nov 25, 2019, 12:14 PM
Authors :
Takumi Mitsuhashi, Juntendo University; Hidenori Sugano, Juntendo University; Madoka Nakajima, Juntendo University; Yasushi Iimura, Juntendo University; Hiroharu Suzuki, Juntendo University; Kaito Kawamura, Juntendo University; Hajime Arai, Juntendo Unive
Rationale: Passive functional mapping (PFM) has been reported as a short time and less invasive functional mapping method. Compared to electrical cortical stimulation mapping (ECSM), sensitivity and specificity of PFM are still inferior and need improvement. In this study, we combined PFM and Cortico-cortical evoked potential (CCEP) to improve functional localization of language. Methods: For this study we reviewed our results of PFM and CCEP for language mapping in 6 epilepsy surgery candidates. Their epileptic foci were supposed to locate in the temporo-parietal area of the dominant hemisphere, and they had subdural electrode implantation for focus localization and language area mapping.We recorded electrocorticography at 1200Hz to 2000Hz sampling rate during PFM. The following language tasks were applied for language mapping; story-listening, word-reading, picture-naming, sentence-reading and question-and-answer. Language-associated areas were detected by the increase of band power more than 25% at 60-90Hz during each task compared to activity at rest.We also carried out ECSMs by continuous stimulation for 2-3sec with 6-12mA at anatomically identified language-associated areas. Based on the results of ECSM, sensitivity, specificity and accuracy of PFM were calculated. CCEPs with alternating stimulus polarity, 0.3ms pulse duration, 1Hz frequency, 6-8mA intensity, and averaged by 50 responses were performed at PFM positive areas. We calculated root mean square (RMS) of 10-100msec of the CCEP response to evaluate functional connectivity between PFM positive areas. Results: Sentence-reading overt task was the best indicator for language mapping with sensitivity: 77.7%, specificity: 70.0%, accuracy: 75.0% and AUC of ROC curve: 0.784. Areas that showed positive by both PFM and ECSM demonstrated high RMS to other language-associated areas. PFM positive and ECSM negative areas showed lower RMS. Conclusions: One of the problems of PFM is the demonstration of wider positive area in comparison to ECSM. Analyzing RMS in CCEP could distinguish false positive areas demonstrated by PFM. Therefore, combination of PFM and CCEP could improve high-gamma functional mapping for language. Funding: No funding
Neurophysiology