Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks

Functional localization of the language and the multiple demand network in individuals with chronic post-stroke aphasia

Poster C61 in Poster Session C, Wednesday, October 25, 10:15 am - 12:00 pm CEST, Espace Vieux-Port

Pieter De Clercq¹, Alicia Gonsalves¹, Tom Francart¹, Robin Gerrits², Maaike Vandermosten¹; ¹KU Leuven, ²University of Ghent

Introduction Aphasia is an acquired language disorder impairing communication ability and is principally caused by a stroke in the language-dominant left hemisphere. Neuronal changes occur in individuals with aphasia (IWA) as they try to restore their language abilities. However, there is ongoing debate about the role of the domain-general multiple demand (MD) network in supporting language recovery. One possible reason for the contradictory findings is the use of group-based region of interest (ROI) analyses that do not account for individual variability in functional networks (Fedorenko et al., 2010). In contrast to previous studies, we aim to determine the language and MD network in a subject-specific manner using functional localization. Additionally, we compare two receptive language localizer tasks, namely a visual (reading) and auditory (spoken language) task, which are believed to represent the same language network in healthy young participants but has not yet been investigated in elderly or IWA. Methods We conducted functional MRI (fMRI) tests on 15 IWA (mean age=64 +- 14 years) in the chronic stage of stroke recovery (>1 year) and 12 healthy controls (HC, mean age=67 +- 9 years). IWA had suffered a stroke in the left hemisphere or bilaterally. Participants performed a spatial working memory task, which included a hard and an easy condition. To suit our target population (i.e., older and stroke population), this task was simplified from Fedorenko et al. (2010), who showed that in young adults the hard>easy contrast robustly activated the MD network. Participants also performed two existing language localizers: a visual reading (sentences>nonwords) and an auditory spoken language (intact>degraded, speech-shaped noise) task, which aim to functionally localize the language network. The method used for network identification followed the approach described by Fedorenko et al. (2010). To investigate the involvement of the MD network during language comprehension, we conducted t-tests (FDR-corrected) to test for activation of single-subject-defined MD regions during the language localizer tasks. Finally, we assessed the spatial overlap between the three functional networks (MD network and language networks defined by visual and auditory localizers) using Dice similarity coefficients. Results During the language reading task, neither group showed significant activation within the single-subject-defined MD network. For the listening task, significant activation was observed in the right (t=5.43, p=0.004) and left insula (t=4.06, p=0.01) of the HC group, but not in IWA The Dice score for the overlap between the MD network and the language networks defined by visual and auditory localizers was 0 for both groups. The mean Dice scores for the two language localizers were 0.27 (SD=0.12) for IWA and 0.23 (SD=0.09) for HC across brain regions. Discussion The MD network does not support language functioning in chronic stroke patients with aphasia or in healthy elderly individuals. The role of the MD network in acute or subacute stroke patients must be explored in the future. While there was some overlap between the language network defined by the auditory and visual localizers, these localizers exhibited differences in activation patterns across language-related brain regions, which will be discussed and further investigated.

Topic Areas: Disorders: Acquired,