You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster C78, Thursday, November 9, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

High gamma neural responses dissociate between the acoustic and linguistic analysis of temporal speech structure

Gregory Cogan1, John Pearson2, Michael Haglund1, Saurabh Sinha1, Tobias Overath2;1Duke University School of Medicine, 2Duke University

Speech perception entails the mapping of the acoustic waveform to its linguistic representation. For this mapping to succeed, the speech signal needs to be tracked across a large temporal range at high temporal precision in order to decode linguistic units (e.g. phonemes, syllables, words). Here we test how cortical processing of such temporal speech structure is modulated by higher-order linguistic analysis. To control the temporal scale of analysis, we used a novel sound-quilting algorithm that controls acoustic structure at different temporal scales; using fMRI, we recently showed that activity in human superior temporal sulcus (STS) increases as a function of temporal scale in an unfamiliar language (Overath et al., 2015). To control the linguistic content, we constructed speech quilts from both familiar and foreign languages. This ensures that any changes at the signal-acoustics level affect both languages identically, while manipulating the linguistic percept differently. Thus, neural responses that vary as a function of segment length but are shared or similar across the two languages suggest analysis at the signal-acoustics level, whereas neural responses that differ based on language familiarity imply the presence of linguistic processing. Here, we recorded electrocorticography (ECoG) from electrodes placed over left temporal or fronto-temporal lobes in three patients who were undergoing pre-surgical monitoring for pharmacologically resistant epilepsy. Patients listened to 6 s long English or Korean speech, quilted with 30 ms or 960 ms segment lengths. Electrodes with significant auditory responses were initially assessed via a permutation test between the 1 s time window following sound onset compared to the pre-stimulus baseline. Neural signals were filtered between 70 and 150 Hz (high gamma) and the results were Bonferroni corrected for multiple comparisons across electrodes. 20/124 electrodes demonstrated a significant auditory response. Within these electrodes, a follow-up analysis showed that sustained high gamma responses throughout the 6 s sounds showed a main effect of segment length (30 vs. 960 ms) in 85% (17/20), a main effect of Language (English, Korean) in 60% (12/20), and an interaction in 50% of auditory electrodes (10/20). Specifically, electrodes that showed an interaction generally displayed a larger increase in high gamma power as a function of segment length in English than in Korean. These results suggest that high gamma neural responses are a potential neural mechanism for tracking speech-specific temporal structure. Ref: Overath T, McDermott JH, Zarate JM, Poeppel D (2015). Nat Neurosci 18:903-911.

Topic Area: Perception: Speech Perception and Audiovisual Integration

Back to Poster Schedule