Abstracts

Rationale: Broca's area plays an important role in different stages of word production, i.e. phonological processing, generating the articulatory code, and coordinating with motor cortex for vocal articulation [1]. However, the exact role of Broca's area in these operations is unclear. High gamma (>50 Hz) electrocorticographic (hgECoG) recording offers a unique opportunity for investigating the role of Broca's area in speech production. In this preliminary study, we investigated hgECoG recordings during six language tasks to study the role of Broca's area during the operations of speech production and perception. Methods: A right-handed 40-year-old male who underwent Phase II epilepsy evaluation participated in this study. The patient was left-hemisphere dominant for language based on the Wada test. Subdural electrodes were placed over the left fronto-temporal regions (Fig. 1). ECoG was recorded while the patient performed 6 language tasks: auditory monosyllable repetition ("ba" or "pa"), overt object naming, overt verb generation (written and spoken nouns), auditory descriptive naming, and auditory word recognition. Time-frequency analysis was performed on ECoG recordings to find significant high gamma activity [2]. Based on the results of cortical stimulation mapping and significant hgECoG activity during language tasks, three electrodes were selected to represent Broca's area (within pars opercularis), the motor cortex, and the superior temporal gyrus (STG). Results: Enhancement of high gamma activity, compared to resting baseline, during 6 language tasks are shown in Figs. 1 and 2. During monosyllable repetition, cortical activation exhibited a temporal progression from STG to Broca's area, and eventually reaching premotor and motor cortex. Activation in Broca's area ended by the time speech begun, though the motor cortex was active before and during speech production. This profile of activation in Broca's area and motor cortex was also observed during the object naming and auditory descriptive naming tasks. During the verb generation tasks (both written and spoken nouns), activation in Broca's area started at ~250 ms following stimulus onset and continued during the early stage of articulation. We observed that Broca's area was not completely inactive during the verb generation task by the time speech commenced. Another interesting observation was activation of the Broca's area during the auditory word recognition task, which is a receptive language task. Conclusions: We used multiple levels of stimulus complexity and different linguistic modalities and found that Broca's area mediates a temporal cascade of neural activity that proceeds from sensory processing to speech production in the motor cortex. Contrary to the classic view of the role of Broca's area in speech production, Broca's area was surprisingly silent by the time speech begun (during some expressive language tasks), though the motor cortex was activated before and during speech production. However, this phenomenon was not observed during verb generation task. Our preliminary data suggest this hgECoG is a robust way to investigate speech function, which we will valid in further subjects. Funding: This study was funded by the Children's Foundation Research Institute, Memphis, TN.