Search Abstracts | Symposia | Slide Sessions | Poster Sessions
A Whole-Brain Model Based on Coupled Oscillators For the Listening Brain?
Poster Session F, Friday, October 2, 2:45 - 4:45 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Hugo Weissbart1, Lucia Amoruso1, Manuel Carreiras1; 1Basque Centre on Cognition, Brain and Language
Whole-brain dynamical models provide a mechanistic link between structural connectivity and measured functional dynamics, as computed from magnetoencephalography (MEG). The Stuart-Landau (Hopf) model, which describes each brain region as a non-linear coupled oscillator, has been extensively validated for resting-state MEG functional connectivity (FC). Concurrently, speech neuroscience has established that cortical oscillations track the rhythmic structure of continuous speech. Here, we present a computational framework that bridges this gap by driving a whole-brain model instantiated as a coupled oscillator network with speech envelope input, while computing frequency-resolved FC metrics. Leveraging recent findings on phase-amplitude coupling between regions in response to acoustic and linguistic features of speech, we further extend this framework by triggering modulation at word onsets. The model successfully reproduces known resting-state dynamics and demonstrates sensitivity to speech-driven modulation of structure-function coupling and global synchronization. Critically, this work aligns with emerging perspectives on the combined roles of evoked potentials and intrinsic oscillations in speech processing. By integrating our network to simulate evoked-potential-like responses—via bifurcation control—we establish a biologically plausible computational framework for how external speech stimuli dynamically shape ongoing neural oscillations. While this framework provides a testable foundation for understanding language processing at an algorithmic level and offers potential heuristics for representing linguistic features, it represents a preliminary yet promising step toward unifying diverse analytical approaches and fields.
Topic Areas: Speech Perception, Computational Approaches