Search Abstracts | Symposia | Slide Sessions | Poster Sessions
High-Gamma Language Mapping Using Stereo-EEG and Complex Linguistic Approach
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Maria Prokopeva1, Sophia Shumilova1, Zhanett Kurtanidze2, Nikita Utyashev3, Andrey Zuev3, Olga Dragoy1; 1HSE University, Moscow, Russia, 2Sechenov University, Moscow, Russia, 3National Medical and Surgical Center Named After N.I. Pirogov, Moscow, Russia
High-gamma mapping (HGM) has been developed as an alternative to the direct electrical stimulation (DES) mapping that is associated with greater risk of provoking seizures. Despite the encouraging results in previous studies showing high sensitivity of the method for pre-/intraoperative language mapping, variability of language tasks and analyses pipelines makes it difficult to compare and generalize the findings. Furthermore, precise assessment of individual variability in neural organization should include a diversity of brain states induced by different tasks. Thus, measuring activation patterns in diverse language contexts may promote a more sophisticated representation of neural correlates of language. We introduce a study of functional language mapping using a battery of language-engaging tasks to assess activation patterns elicited by different types of linguistic processing. Eleven patients (4 males, 1 left-handed, M age=27.1, SD age=10.9, age range: 10–43 years) with drug-resistant epilepsy took part in this research. Before stereo-EEG implantation each patient underwent s- and dw-MRI recording, followed by manual language tracts reconstruction. After the electrodes implantation patients performed three language tasks: visual object naming, auditory pseudoword repetition, and auditory sentence completion. For the analysis we notch-filtered (50 Hz) the data, performed Laplacian re-referencing, epoched data with respect to the stimulus onset, computed power-spectral density for the high gamma (70-150 Hz) and alpha-beta bands (8-25 Hz). The electrodes were identified as “active” if there was a significant increase in high-gamma power accompanied by a decrease in the low frequencies during task performance as compared to the baseline period. The electrodes were labeled on post-implantation sMRI and coregistered to reconstructed tracts in order to determine their anatomical distribution. The majority of active electrodes were located in the left hemisphere (60% and 17% in the grey and white matter, respectively). Object naming and pseudoword repetition most commonly produced language-related activation detected in the study. Interestingly, there was a noticeable dissociation between these two tasks. Whereas object naming was associated with the occipital and middle to posterior temporal cortices, pseudoword repetition involved the anterior temporal and inferior parietal cortex. The intersection between both tasks was evident in the inferior parietal lobule. Relatively few cortical areas were associated with the sentence completion task exclusively. The same electrodes were additionally involved in object naming. This intersection could reflect signals related to lexical access required for both aforementioned tasks, but not pseudoword repetition, that lacks lexical clue. Intersection of all three tasks was found mainly in the pre-/postcentral gyri, the superior temporal cortex, and the cortico-spinal tract, which could imply that primary auditory and motor components are largely the only language aspects that they share. Tracts reconstruction additionally revealed that the active electrodes in the white matter crossed main language-related fasciculi (i.e., arcuate fasciculus, inferior fronto-occipital and inferior longitudinal fasciculi), although the association between the particular tracts and tests specific to them was not evident and will be further investigated. While HGM is a promising tool for passive functional mapping, validation against DES is required to establish sensitivity and specificity of the method.
Topic Areas: Disorders: Acquired, Methods