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

Characterization of functional connectivity in post stroke aphasia during fMRI naturalistic and language tasks

Poster B61 in Poster Session B, Tuesday, October 24, 3:30 - 5:15 pm CEST, Espace Vieux-Port
This poster is part of the Sandbox Series.

Nicole Carvalho1, Anne Billot1, Maria Varkanitsa1, Isaac Falconer1, Niharika Jhingan2, Swathi Kiran1; 1Center for Brain Recovery, Boston University, Boston, MA, 2Brain & Cognitive Sciences, Massachusetts Institute of Technology, Boston, MA

Language network function is able to be reliably measured in healthy controls using language localizer tasks (Fedorenko et al., 2010). These language localizer tasks involve the reading of words and nonwords, presented in blocks to localize the language network in each subject1. Yet these language localizer tasks are constrained in timing and do not capture natural language processing. During naturalistic language tasks, predictive coding was found in the language network of healthy controls (Shain et al., 2020) . Yet the functional connectivity of naturalistic language processing has yet to be characterized in post stroke aphasia. This is critical for understanding how language is organized in this population. Resting state functional connectivity has also been found to be disrupted in post stroke aphasia (Klingbiel et al., 2019). In the current study, we seek to characterize differences in functional connectivity in people with post stroke aphasia and healthy controls across different states (rest, naturalistic language processing, and language localizer task). Magnetic resonance images (MRIs) are being collected from 40 individuals with chronic aphasia (> 6 months) due to a left hemisphere stroke and 40 healthy controls. They will undergo an MRI collected on a Siemens 3T Magnetom Prisma. During the scan, subjects will complete a resting state scan, a naturalistic language task and a language localizer task. The naturalistic language task consists of story listening. A five minute long story is presented via earbuds and subjects are shown a fixation cross throughout. Following the story, yes no questions are asked to ensure comprehension. The language localizer task consists of reading sentences and lists of nonwords in a block design. We plan to conduct a joint independent component analysis (jICA) using the three conditions from each subject as input. jICA is preferred over a traditional independent component analysis (ICA) because the jICA takes the lesion into account when producing component masks where traditional ICAs only mask the lesion. Then the resulting components will be used as regions of interest in a functional connectivity analysis within each condition. Then we will compare the resulting connectivity matrices across the two subject groups and three conditions. Across the three conditions, we expect to see a spectrum, with the language localizer condition being the most language like, and the resting state being the least language like. Across the two groups, we expect to see the most similar connectivity patterns in the resting state condition and the least similar connectivity patterns in the language localizer condition, with the naturalistic language condition falling somewhere between the two. Preliminary data showed increased connectivity in certain components during story listening when compared to resting state. In turn, these same components had increased connectivity during the language localizer task when compared to story listening.

Topic Areas: Disorders: Acquired,

SNL Account Login

Forgot Password?
Create an Account