Meeting Banner
Abstract #3748

The LoCo: A Measure of Gray Matter Structural Connectivity Loss and Its Application to Neurodegenerative Disorders

MAGNA25Amy Kuceyeski1, Yu Zhang2, 3, Ashish Raj1

1Radiology, Weill Cornell Medical College, New York, United States; 2Center for Imaging of Neurodegenerative Diseases, Department of Veteran's Affairs Medical Center, San Francisco, CA, United States; 3Radiology, University of California San Francisco, San Francisco, CA, United States

We implement a metric called Comparative Connectivity Loss (CCL) that gives the amount of structural white matter connectivity disruption incurred by a gray matter region for a particular pattern of white matter integrity loss. This metric is calculated on a standard atlas for three groups, Alzheimer's disease, fronto-temporal dementia, and age-matched normal controls. We show significant correlations of CCL with atrophy patterns in the two diseases. In addition, we show that the CCL outperforms gray matter atrophy when classifying individuals into the three groups, while having similar levels of accuracy with white matter integrity measures of radial and longitudinal diffusivity.

Abstract PDF E-Poster Video

Keywords

able addition affairs affected affecting agree agreement anisotropy assessment atlas atrophy better brain called cerebellum cerebral classically classification clear closer cognitively cohort cohorts comparative computational connected connecting connectivity consider constructed control controls cook correct correction correlate correlation correlations cortical cortices cross damage damaged default deficit degeneration dementia dementias differentiating diffusion diffusivity dimensionality disease diseases disorders displays disruption earlier early eigenvalues either exam exploits fiber fibers fractional frontal general generally give gray greater gross healthy highest highly identified individual inferior injured injury integrates integrity involvement journal knowing known lack linear loco longitudinal loss losses many maps matched measurable measure measures measuring mediated medical mental methodology metrics much neurology normalized occipital pathology pathways patients pattern population post posterior progression propagation propose proteins quantitative radiology recorded reduction represent retraction role scores sensitive sensitivity significance spatially statistics structural structures studies subjects synaptic taken temporal thresholded tissue topology tracts treatment unbiased versus veteran volume wise young