Understanding speech in difficult listening situations is challenging. A popular technique to simulate difficult listening situations while still enabling high parametric control is noise-vocoded speech. Recent findings by our group have shown that during periods of continuous vocoded speech alpha power decreases with increased levels of degrading, whilst speech-brain coherence (1-7 Hz) only increased as long as speech was still comprehensible. This putatively captures the underlying compensatory processes following hearing loss and cochlear implant surgery. However, different processes may be recruited when listening is only temporarily impaired. In these situations information obtained during periods of clear speech might give important context information for periods where speech is not comprehensible. Here, we investigated how intermittent degradation of an otherwise clear speech stream affects speech processing. We used several measures to capture both the spectral and temporal dynamics of neural speech processing at cortical and subcortical levels of the auditory hierarchy using MEG. We observed (as previously reported) that alpha power declines with a decrease in intelligibility. However, opposed to previous findings speech-brain coherence decreases as well. This pattern was already starting to emerge at putatively subcortical processing stages. While the speech-brain coherence decreased with intelligibility we noted a more precise tuning in the coherence spectrum to the syllable rate of speech (4-5 Hz) when speech was vocoded (7-Channels). In sum, we show that the activation patterns involved in the processing of vocoded speech strongly depend on whether clear speech offers a context for the vocoded audio stream or not.
|Online-Conference||APAN 2020 - Advances and Perspectives in Auditory Neuroscience|
|Abbreviated title||APAN 2020|
|Period||22/10/20 → 23/10/20|
- brain processing of speech and language
- subcortical auditory processing