Search Papers | Poster Sessions | All Posters
Poster C65 in Poster Session C - Friday, August 9, 2024, 11:15 am – 1:15 pm, Johnson Ice Rink
Human cortical auditory sensitivity adaptation to simulated hearing loss: A data-driven fMRI approach with naturalistic speech
Arkan Al-Zubaidi1 (), Jochem Rieger1; 1Applied Neurocognitive Psychology Lab and Cluster of Excellence Hearing4all, University of Oldenburg
The naturalistic auditory stimulus reveals dynamic changes in cortical temporal lobe activity as participants engage with continuous speech (Boos et al., 2021). However, there is little knowledge regarding cortical auditory speech processing in response to clear and degraded naturalistic stimuli. In this fMRI study, we used voxel-wise spectrotemporal receptive field (VWSTRF) estimation to assess the neuronal population sensitivity to acoustic features (mel spectrogram [MS]) in different auditory areas for clear stimuli and at two levels of simulated hearing loss (SHL). The audio description of the movie "Forrest Gump" soundtrack was employed as a clear naturalistic stimulus. VWSTRF models with MS features predicted the BOLD responses in the Heschl gyrus and planum temporale. We find that the spectrotemporal sensitivities change in a non-trivial way. The neuronal population sensitivities increased with SHL in low-frequencies (100-1000 Hz), where stimulus degradation was milder, and decreased in high frequencies (4-8 kHz), where stimulus degradation was heaver. This suggests a compensatory sensitivity increase in the frequency range below 1kHz, which is relevant for speech comprehension. Conversely, the observed sensitivity reduction in the high-frequency range (>4 kHz) aligns with the assumption that the neuronal population implements passive filtering in this frequency range, which is less relevant for speech comprehension. In conclusion, the present study demonstrates that the spectrotemporal sensitivities of neural ensembles exhibited rapid, potentially task-relevant changes while participants listened to continuous speech in natural environments with different levels of SHL.
Keywords: simulated hearing loss naturalistic speech voxel-wise spectrotemporal receptive field model spectrotemporal sensitivities