DYNAMICAL MAPPING OF CORTICAL LANGUAGE NETWORKS BY HIGH-GAMMA EVENT-RELATED CAUSALITY (ERC)
Abstract number :
A.02
Submission category :
3. Clinical Neurophysiology
Year :
2008
Submission ID :
9018
Source :
www.aesnet.org
Presentation date :
12/5/2008 12:00:00 AM
Published date :
Dec 4, 2008, 06:00 AM
Authors :
Anna Korzeniewska, Piotr Franaszczuk, R. Kus, C. Crainiceanu and Nathan Crone
Rationale: Because electrocortical stimulation mapping (ESM) may cause excessive afterdischarges or seizures, passive electrocorticographic (ECoG) recordings have been explored as a potential means of functional mapping in patients undergoing epilepsy surgery. These studies have shown that functional activation of motor and/or language cortices is associated with an event-related increase in high-gamma (~60-100 Hz) ECoG activity that appears to be more specific than other spectral measures of cortical activation, including activity in and around 40 Hz. However, little is known about the role that high-gamma activity plays in large-scale cortical network interactions, which are presumed to be necessary for language and other higher cortical functions. Methods: ECoG was recorded during two word production tasks (object naming and word repetition) that used two stimulus modalities (auditory vs. visual) and two response modalities (spoken vs. signed). Two methods - Event Related Causality (ERC) and Matching Pursuit (MP) were used to analyze event-related changes in ECoG high-gamma (70-115 Hz) energy and in causal interactions between cortical sites at high-gamma frequencies, respectively, in cortical language networks. The results were compared under the different language task conditions,and the most active sites revealed by ERC analysis were compared to those revealed by MP analysis, and to clinical electrical mapping. Results: Event-related increases in ECoG causal interactions within Broca's area and between this area and other studied regions were observed for all language tasks. During auditory perception of spoken words, interactions with superior temporal gyrus predominated, whereas visual perception of pictures of objects evoked interactions with occipital-temporal cortex. Spoken responses were associated with interactions involving mouth/tongue motor cortex, while interactions during signed responses involved hand motor cortex. Sites with predominantly incoming high-gamma activity flows, shown by ERC, coincided with sites having maximal event-related increases in high-gamma energy, shown by MP. There was also good concordance between these ECoG results and the results of clinical electrical mapping. Conclusions: Word production tasks requiring dynamic cortical network interactions are associated with event-related changes in the flow of high-gamma activity (70-115 Hz) between and within cortical areas likely mediating different components of these tasks. The topographic patterns of interactions between different cortical areas participating in word production depend on response modality (speech vs. sign language), as well as on stimulus modality (visual vs. auditory). Large and numerous inflows to a ECoG recording site (according to ERC) are associated with event-related ECoG energy increases (according to MP) at the same site, at the same high-gamma frequencies. Changes in energy and/or causal network interactions in high-gamma frequencies may help localize cortical areas critical for language performance. Supported by: NINDS R01 NS40596
Neurophysiology