You are viewing the SNL 2018 Archive Website. For the latest information, see the Current Website.

Poster B40, Thursday, August 16, 3:05 – 4:50 pm, Room 2000AB

Functional modularity supports treatment-induced recovery in chronic aphasia

Yuan Tao1, Brenda Rapp1;1Johns Hopkins University

Network analyses of functional connectivity have revealed that the brain is organized in a non-random, modular structure consisting of densely inter-connected clusters (or modules) of brain regions. Critically, modular structure relies on two types of “hubs”: “local hubs” that connect regions within one module, and “global hubs” that facilitate communication between modules. Modular structure in functional connectivity has been linked to cognitive performance in healthy populations (e.g. Bassett et al., 2011) and various neurological disorders (e.g. Bucker et al., 2009; Lynall et al., 2010; Duncan & Small, 2016). Understanding how lesions affect modularity structure can provide critical insights into neural plasticity. Computer simulation research predicts that lesions affecting hub areas will have the most widespread effects (Honey & Sporns, 2008). However, the consequences of actual lesions and how they relate to behavior is less clear. In the current study, we evaluated the modular structure in participants with acquired dysgraphia (n=15, 4 females, age 61+/-10) resulting from a left-hemisphere stroke (>1 year post-stroke), and examined the change of the modularity measure (Newman, 2006) in relation to intensive behavioral treatment for dysgraphia. FMRI data were collected while participants performed a spelling task before and after the intervention. We estimated the whole-brain functional connectivity (lesioned regions excluded) for each participant at each time-point, and calculated modularity based on a modular structure derived from an age-matched, healthy control group performing the same scanner task. Results We found a significant increase in modularity from pre- to post-treatment (p<0.01). Regression analysis demonstrated that higher modularity scores before treatment were related to less severe deficits (p<0.05) and greater treatment gains (p<0.1), suggesting that higher modularity scores index a system with higher functionality which might, in turn, retain greater re-learning capacity. The effects were similar in both the ipsilesional (LH) and contralesional hemispheres (RH). To investigate the underlying bases for the observed modularity increases, we identified the global and local hubs from the control data and examined their connectivity properties in the lesioned brains. The global hubs (i.e., inter-module connectors) exhibited no differences between the lesioned and the control group, neither at pre nor post-treatment. In contrast, the local hubs (i.e., within-module connectors) exhibited lower within-module density in the lesioned participants and, furthermore, these values increased significantly from pre to post-treatment. These results indicate that the lesions primarily affected the within-module connectivity that was strengthened by the treatment, supporting the observed improvements in spelling. Conclusion We used modularity, a graph-theoretic measure, to assess the neural integrity of the functional connectivity of participants with chronic stroke-induced dysgraphia before and after behavioral treatment. We found that modularity indexed deficit severity and extent of re-learning such that an increase in modularity was associated with behavioral improvement. Consistent with other network-based connectivity studies (e.g. Gratton et al., 2012), our findings demonstrate that modular structure is an important organizing principle of the brain and that modularity can be a useful tool for evaluating post-stroke neural re-organization.

Topic Area: Language Disorders