Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Functional connectivity patterns in post stroke aphasia: task level differences
Poster Session B, Friday, October 25, 10:00 - 11:30 am, Great Hall 3 and 4
Nicole Carvalho1, Maria Varkanitsa1, Isaac Falconer1, Anne Billot2,3, Swathi Kiran1; 1Center for Brain Recovery - Boston University, 2Department of Neurology, Massachusetts General Hospital & Harvard University, 3Department of Psychology, Center for Brain Science, Harvard University
Introduction Traditionally, the study of the neural correlates of language processing in people with aphasia (PWA) has relied on language-constrained tasks to engage language-related brain regions and understand the functional organization of the language network. In addition, resting-state fMRI has also been used to measure altered connectivity patterns in PWA. Tasks, such as resting-state and language localizers, are constrained in timing and/or do not fully capture the nuances of naturalistic language processing. The aim of this study is to determine whether functional connectivity (FC) patterns differ in tasks that have varying language demands. Methods 29 chronic PWA (4F, age: mean=58.7 years, range=19.7–75.7 years, time post stroke: mean=90.6 months, range=5.2–530.1 months, WAB-AQ: mean=80.8, range=31.5-99.6) due to left hemisphere stroke lesion and 27 neurotypicals (21F, age: mean=64.3 years, range=44–80 years) participated in the experiment. All subjects completed a structural and functional MRI protocol including resting-state, movie watching (e.g., a silent animated short PIXAR film), story listening (e.g., a simple short story), and language localizer (e.g., sentences and lists of nonwords presented one word or nonword at a time, in a blocked design) tasks, lasting about 5 minutes each. Lesions were masked out using ITK-SNAP. Data was preprocessed using the CONN EL pipeline (https://web.conn-toolbox.org/resources/conn-extensions/evlab), including segmentation, normalization, motion correction, co-registration, resampling, removal of outliers beyond default CONN thresholds, and smoothing. Then, a region of interest (ROI) to ROI connectivity analysis was conducted using the default CONN atlas. Mean composite whole-brain and within-network connectivity estimates were calculated, resulting in nine estimates for each participant (i.e., whole-brain and cerebellar, dorsal attention (DAN), default mode (DMN), frontoparietal, language, salience, sensorimotor, and visual networks) per task. These estimates were entered as dependent variables in regression models looking at the group, task, and interaction effects between group and task. Results Overall, the main group effect showed that PWA exhibited significant lower whole-brain and within-network FC than neurotypicals. At the whole-brain level across groups, the language localizer task elicited higher connectivity compared to rest (p=0.001), movie watching (p=0.0001) and story listening (p<.0001). Within-network analyses showed (a) no differences in connectivity patterns across tasks in the language, frontoparietal, salience, sensorimotor, and cerebellar networks, (b) overall, story listening elicited significantly lower connectivity than the language localizer task in the DMN (p=0.001), significant interaction effects between group and task in the (c) DAN and (d) visual network between all tasks except the language localizer. During story listening (p=0.0008), movie watching (p<.0001), and resting-state (p<.0001), PWA exhibited significantly lower connectivity than neurotypicals in the DAN. During story listening (p<.0001), movie watching (p=.0368) and resting-state (p=.0001), PWA exhibited significantly lower connectivity than neurotypicals in the visual network. Conclusions In line with previous research, results indicate that overall PWA exhibited lower connectivity than neurotypicals. Language tasks elicited similar FC patterns in the language network, yet they elicited varying levels of connectivity in non-language networks. This variation of FC patterns may be a function of language processing that may be indicative of a PWA’s deficit or may serve to support language processing in a lesioned brain.
Topic Areas: Disorders: Acquired,