An Investigation of the Neural Dynamics Underlying Speech Perception and Production: EEG Evidence for Internal Inverse and Forward Models
Year: 2024
Type: dissertation
Abstract: Dual stream models of language processing provide a functional neuroanatomical model for auditory speech processing, however, the neural dynamics underlying speech processing as well as the sensorimotor transformations in these pathways are still poorly understood.Internal models predict the interaction between sensory and articulatory representations: forward internal models predict sensory outcomes of speech production, linking motor areas to perceptual areas; inverse internal models, on the other hand, link speech outcomes to motor plans by bridging sensory to articulatory representations during speech perception.In this study, I used electroencephalography (EEG), carefully designed behavioral stimuli, and multivariate analysis techniques to systematically investigate the nature of neural representations and their predicted dynamic interactions during speech perception and production.First, I validated the approach by replicating prior electrocorticography (ECoG) findings regarding the latency and organization of speech information during perception.I then found that the sensorimotor transformation between different neural representations during speech perception and production corresponded to the hypothesized internal model framework.Specifically, I found evidence for an inverse model from perceptual representations in temporal channels to articulatory representations in superior frontal channels during speech perception, and for a predictive forward model from frontal to temporal iv channels active during speech production.Intriguingly, this forward model appeared to include the anteroventral stream.In addition, I found evidence that the organization of speech representations in different parts of the brain's speech network is task-dependent and dynamic.
Author: Nikolov, Plamen Petrov
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W6991881001