Abstract Details

Title Impaired Global Network Topology in Residual Connectomes in Stroke-induced Aphasia

Topic Aging, Dementia, and Behavioral Neurology

Presentation(s) P5 - Poster Session 5 (5:30 PM-6:30 PM)

Poster/Presentation
Number 9-005

Objective To examine global network organization among spared brain regions in stroke-induced aphasia.

Background Most studies of stroke-induced aphasia examine the relationship of lesioned brain regions and white matter tracts and their association with symptoms. However, the configuration of spared edges and brain networks has been relatively unexamined. Further, because longitudinal comparisons are difficult to conduct to control for neuroplasticity post-injury, cross-sectional approaches that support inferences accounting for possible neural remodeling are necessary.

Design/Methods In 39 subjects with chronic stroke-induced aphasia and 39 age-matched controls, we applied a novel anatomy imputation procedure to obtain comparable anatomical T1 imaging parcellations in health and stroke. Then, we reconstructed white matter fractional anisotropy (FA) networks on streamlines using anatomically constrained diffusion tractography in 234 regions from the Lausanne parcellation. In a “simulated attack” analysis, we created personalized null distributions of global network statistics (small world propensity, global efficiency, transitivity, and density) that would be expected if a given stroke represented in each real lesion map had been observed in each of the controls. Then, we computed the percentile of each of the network statistics to determine whether they were impaired or enhanced relative to their personal null.

Results While small-world propensity and density were typically (>90% of subjects) well within the null distributions, global efficiency and transitivity were often reduced (>75% of subjects). These findings indicate that post-injury symptoms could result from reduced local and global information processing above and beyond gross losses in connectivity

Conclusions Simulated attack to healthy FA networks can support sensitive analyses of impaired network topology in residual stroke networks. The contribution of stroke risk factors and secondary effects of stroke to topological disruption should be specifically examined. Further, the validity of simulated attack measures applied to control networks and their behavioral relevance should be explored.

Authors/Disclosures
John D. Medaglia, PhD (Drexel University) PRESENTER	The institution of Dr. Medaglia has received research support from National Institutes of Health.
	No disclosure on file
	No disclosure on file
Peter Turkeltaub, MD, PhD (Georgetown University)	Dr. Turkeltaub has nothing to disclose.

��ɫ��

��ɫ��