Abstract
OBJECTIVE: To determine the optimal montage and vocalization conditions for high-gamma language mapping using stereoelectroencephalography. METHODS: We studied 12 epilepsy patients who underwent invasive monitoring with depth electrodes and measurement of auditory-naming related high-gamma modulations. We determined the effects of electrode montage and vocalization conditions of the response on the high-gamma (60-140 Hz) amplitudes. RESULTS: Compared to common average reference montage, bipolar and Laplacian montages effectively reduced the degree of auditory naming-related signal deflections in the white matter during the stimulus and response phases (mixed model estimate: -21.2 to -85.4%; p < 0.001), while maintaining those at the cortical level (-4.4 to +7.8%; p = 0.614 to 0.085). They also reduced signal deflections outside the brain parenchyma during the response phase (-90.6 to -91.2%; p < 0.001). Covert responses reduced signal deflections outside the brain parenchyma during the response phase (-17.0%; p = 0.010). CONCLUSION: On depth electrode recording, bipolar and Laplacian montages are suitable for measuring auditory naming-related high-gamma modulations in gray matter. The covert response may highlight the gray matter activity. SIGNIFICANCE: This study helps establish the practical guidelines for high-gamma language mapping using stereoelectroencephalography.