Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Language assessment during awake brain tumor surgery and ECoG
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Patricia Silva de Camargo1, Adam, Morgan2,3, Giovanna Souza3, Leonidas Angelin3, Analía Arévalo3, Guilherme Lepski3,4; 1Institute of Biosciences, University of São Paulo, 2Grossman School of Medicine, New York University, 3Medical School, University of São Paulo, 4Department of Neurosurgery, Eberhard Karls University
Functional language assessment during awake craniotomy is critical for preserving eloquent cortex while maximizing tumor resection. To address this, we developed a comprehensive language battery (originally created in Brazilian Portuguese) integrated with a value-based decision scale to guide resection of language-related areas. This full language battery comprises different tasks: object naming, action naming, verb generation, odd-word-out, word repetition, pseudoword repetition, auditory naming, and verbal fluency tasks (semantic and phonemic), which are not commonly incorporated into standard intraoperative language protocols. This approach enables fine-grained mapping of both canonical and non-canonical language functions intraoperatively. Intraoperative electrocorticography (ECoG) complements this approach by enhancing language mapping accuracy, reducing surgical time, and minimizing postoperative aphasia risk. We present a case study of a 30-year-old woman with a low-grade glioma and a history of seizures undergoing resection of a large temporoparietal tumor. Surgery was performed under local anesthesia to optimize language monitoring and reduce the chance of seizures. A 128-channel ECoG grid was used to record language-related neural activity during the language battery (excluding verbal fluency tasks), without electrical stimulation. Verbal fluency tasks were applied before, during, and 5-months after surgery. These tasks revealed a marked impairment, in stark contrast to performance on all other tasks. Before surgery, the patient, who was a biologist, produced unusually low-frequency examples during semantic fluency (e.g., “pufferfish” instead of common responses such as “cat”), suggesting atypical lexical retrieval strategies. During surgery, verbal fluency performance declined markedly over six tests (words generated per test: 8-9-9-2-0-0), indicating significant disruption of language processing. Based on this marked decline, part of the tumor was intentionally preserved to avoid damage to the eloquent cortex. Five months later, verbal fluency improved to approximately ten items per category, with recovery of more typical high-frequency responses. A preprocessing and analysis pipeline targeting high gamma activity (70-150 Hz), a well-established marker of local cortical engagement, was developed and applied to the recorded data. While data analysis is ongoing, preliminary results revealed possible task-related high gamma activity in a subset of electrodes overlying the tumor margin.
Topic Areas: Speech Perception, Language Production