Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Language network dysfunction with preserved temporal variability of dynamic functional connectivity in individuals with post-stroke aphasia
There is a Poster PDF for this presentation, but you must be a current member or registered to attend SNL 2023 to view it. Please go to your Account Home page to register.
Poster A52 in Poster Session A, Tuesday, October 24, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Isaac Falconer1, Anne Billot1, Maria Varkanitsa1, Nicole Carvalho1, Niharika Jhingan2, Swathi Kiran1; 1Boston University, 2Massachusetts Institute of Technology
Introduction: Variation in post-stroke aphasia (PSA) recovery and underlying neurobiological mechanisms remain poorly understood. Functional MRI studies have revealed treatment-induced changes in resting state functional connectivity that may represent functional reorganization supporting language recovery. Dynamic functional connectivity (dFC), which remains largely unstudied in the context of PSA, can shed more light on this functional reorganization. We have previously found that greater temporal variability (TV) of language network (LN) dFC (i.e., the magnitude of second to minute timescale variations in inter-regional synchronizations) is associated with (1) greater response to aphasia therapy aimed at improving word-finding and (2) greater treatment-induced increases in LN small-worldness, a graph metric that measures both the efficiency of communication across a network and the tendency of nodes to cluster together. Even though TV appears to be promising towards clinically focused prognosis, it is still not clear if, and to what extent, TV is affected by stroke and how dFC is altered in patients with PSA. The answer to these questions will provide important context for studies relating dFC to PSA recovery and functional reorganization. Methods: We compared dFC in 19 patients with chronic PSA (>6 months post-stroke) to that of 42 healthy controls (HC). Resting state functional MRI scans and T1-weighted structural scans were collected using Siemens Prisma 3T scanners. Lesion maps were generated using the semi-automated segmentation tool ITK-Snap and were used to mask lesioned voxels from functional volumes. All MRI data was preprocessed using SPM12 and the CONN Evlab module. Mean time series per region of interest (ROI) were generated using the CONN toolbox and the Automated Anatomical Labeling (AAL3) atlas and sliding window dFC and TV were computed for 34 bilateral language ROIs using custom MATLAB scripts. K-means clustering was used to identify connectivity states. Fractional occupancy (FO, i.e., the fraction of windows clustered into a state) was computed for each state and each participant. Finally, two-sample t-tests were used to compare TV and FO in the PSA and HC groups. Results: No significant difference in TV was found between PSA and HC groups (PSA: mean=0.56, HC: mean=0.55, p=0.80) indicating that TV is not significantly altered in individuals with PSA. However, the clustering analysis did reveal group differences. Four connectivity states were identified, two of which showed significantly different FO between PSA and HC groups. Individuals with PSA spent more time in a state characterized by relatively low small-worldness (1.27) and global efficiency (3.72e-4) (PSA mean FO=0.39, HC mean FO=0.093, p<0.001). HC spent significantly more time in a state with higher small-worldness (2.0) and global efficiency (5.2e-4) (PSA mean FO=0.15, HC mean FO=0.29, p<0.05). Conclusion: Individuals with PSA were found to have similar TV to healthy controls suggesting that the relationship with treatment response is reflective of potentially modifiable interindividual differences in the capacity for functional reorganization and recovery, rather than stroke-related effects. Individuals with PSA were, however, found to spend more time in a state with less efficient communication and less clustering of sub-networks, likely representing dysfunction in LN due the stroke lesion.
Topic Areas: Disorders: Acquired, Computational Approaches