Schedule of Events | Symposia

Navigating the Neural Landscape of Language Comprehension at Millimeter/Millisecond scale: fMRI-EEG fusion analysis

Poster Session E - Monday, March 31, 2025, 2:30 – 4:30 pm EDT, Back Bay Ballroom/Republic Ballroom

Clair Min Kyung Hong¹ (min.kyung.hong@vanderbilt.edu), Katherine Aboud²; ¹Vanderbilt University

To comprehend written or spoken language, the brain must rapidly coordinate communications across the whole brain to build a coherent internal representation. While decades of work have identified neural signals underlying language comprehension (LC), key limitations in brain imaging have prevented the real-time (millisecond) and whole-brain (millimeter) characterization of LC. We addressed these methodological constraints by using a novel application of fMRI/EEG fused multimodal brain imaging methods in healthy adults (n = 30) to identify the spatio-temporal progression of neural network engagement supporting LC in the one second following comprehension. We identified five key functional brain networks and leveraged text feature analysis to determine each component's sensitivity to specific cognitive dimensions of LC: occipitotemporal perceptual word processing network (associated ERP peak at ~250ms), temporoparietal semantic meaning retrieval network (~400ms), posterior default mode inferential network (~500ms), frontotemporal semantic integration network (~600ms), and a goal-directed comprehension network with default mode and frontoparietal control network nodes (~700ms). Interestingly, posterior DMN activations (~500 ms) related to inferential processing acted as a "hinge point" between early word reading and later higher-order networks: high LC performers showed increased reliance on this bottom-up inferential network, which coincided with decreased reliance on top-down semantic integration areas. These findings suggest that naturalistic LC is characterized by rapid trade-offs between perceptual, core language, and higher-order cognitive networks, and the dynamics between these systems are dependent on expertise. These findings provide insights into the neural mechanisms supporting fluent language processing and offer a methodological framework for investigating related clinical disorders.

Topic Area: LANGUAGE: Other