Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Unraveling the neural correlates of natural language processing and post-stroke language recovery: A comprehensive and lesion-agnostic approach
Poster Session B, Wednesday, September 30, 4:30 - 6:30 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Ella Eycken1, Andi Smet1, Mara Barberis1, Bastiaan Tamm1, Robin Gerrits2, Miryam de Lhoneux1, Maaike Vandermosten1; 1KU Leuven, 2Max Planck Institute for Human Cognitive and Brain Sciences
Introduction: Until today the complex system of language processing in the human brain is not yet fully understood. Traditionally there is a sole focus on the core left-lateralized language system which relates to language microstructure, such as how sounds, meaning, and grammar are processed. However the right hemisphere is also involved in language processing, but its role is unclear. Moreover, natural language comprises more than microstructure and involves also macrolinguistics discourse aspects. Lastly, language cannot simply be associated with localized brain regions and must be interpreted as a circuit of brain connections. This project wants to overcome the gaps introduced by traditional studies by following a holistic, lesion-agnostic approach. More specifically, this project aims to uncover the neural underpinnings of natural language processing, by performing whole-brain analyses of brain regions and connections involved in micro- and macrolinguistic processing in patients with (sub)clinical language difficulties after stroke. Additionally, the neural substrates of language recovery after stroke will be investigated by including longitudinal language assessment over the course of the patients rehabilitation process. Methods: Natural language production will first be assessed in 300 acute left (LH) and right hemispheric (RH) stroke patients, using ecologically valid bedside natural speech tasks that allow to extract multiple micro- and macrostructural language characteristics simultaneously using Natural Language Processing. Next, language abilities are followed up in a subsample of 100 patients (evenly balanced between LH and RH stroke) by means of a monthly telephonic interview that has been developed to efficiently elicit rich semi-spontaneous speech, and that will first be validated in 90 healthy volunteers. Finally, extensive language tests are administered at six months post stroke at the patients homes. Micro- and macrolinguistic characteristics extracted from the speech samples will then be correlated to lesion delineations on clinical MRI-images using advanced lesion- and disconnection-symptom mapping, to reveal whole-brain network-level neural substrates of language processing and recovery after stroke. Expected Results: A voxel-wise mass-univariate lesion-symptom mapping study (Barberis et al., 2026) in a separate set of 49 chronic LH stroke patients with aphasia revealed a significant neural substrate for lexico-semantics in left inferior parietal, superior and middle temporal and lateral occipital regions, based on 16 natural language measures extracted from a picture description task. By extending this approach to multivariate and disconnection-based analyses in 300 acute stroke patients, we expect to find mainly LH regions to be involved in microstructural processing, while distributed networks in both hemispheres would be crucial to macrostructural processing. After finishing ongoing validation analysis of the telephone interviews, we expect longitudinal mapping analyses in 100 stroke patients to differentiate brain structures leading to transient or permanent deficits in micro- and macrolinguistic processing after stroke. Conclusions: By applying lesion- and disconnection-symptom mapping analyses to micro- and macrolinguistic characteristics of natural speech, this project contributes to uncovering the brain regions and connections that critically sustain verbal communication and its recovery after stroke, with attention to the underexplored role of the right hemisphere.
Topic Areas: Language Production, Disorders: Acquired