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

Poster C41, Friday, August 17, 10:30 am – 12:15 pm, Room 2000AB

Diversity of modular networks determines aphasia severity.

Barbara Khalibinzwa Marebwa1, Julius Fridriksson2, Leonardo Bonilha1;1Medical University of South Carolina, 2University of South Carolina

About 30-40% of dominant hemisphere stroke survivors sustain permanent and disabling language problems, and factors that drive incomplete recovery remain unclear. Cognitive function arises from topological interplay that can be captured from the connectome’s community structure, and the topological integrity of residual white matter networks driving language recovery is yet to be fully investigated. We therefore hypothesized that patients with a disorganized and less diverse topology would suffer worse aphasia. Using post-processing methods of diffusion tensor imaging optimized for lesioned brains, we reconstructed the individual structural whole-brain connectome from 83 right handed participants with a single left hemisphere ischemic or hemorrhagic stroke (age 60.1± 9.1 years, 28 females). All participants underwent language assessment using the Western Aphasia Battery that yields a global measure of aphasia severity on a scale of 0-100 (WAB-AQ, mean 68.2±28.7). We determined assortative network organization using Newman’s modularity algorithm. We then calculated a consensus partition that gave the optimal organization of the community structure after 100 different optimizations. Using the maximum number of partitions obtained from each individual, we further fit the weighted stochastic block model (WSBM) to each connectome. The WSBM uncovers different motifs of assortative, dis-assortative, and core-periphery communities. Finally, we calculated the diversity index of each node, which is a measure of how much the node participated in the different community motifs. We found that left hemisphere modularity was significantly associated to the left hemisphere mean diversity score (r = -0.3, p = 0.004), meaning that less fragmented networks were more diverse and had more community motifs. We also found that higher left hemisphere modularity was associated more severe aphasia (r = -0.4, p < 10-4) meaning that fragmentation of left hemisphere networks led to less diversity, poor communication among modules, and more severe aphasia.

Topic Area: Language Disorders