using hitting-time characteristics to detect sensory pathways during resting-state
Name: Paria Rezaeinia
Grad Year: 2019
Graph theory has been used to successfully model information processing in the brain but common implementations do not take advantage of established regularities in brain structure like the hierarchical processing of visual stimuli. We seek to identify brain graphs with structures similar to those expected during the hierarchical processing of visual stimuli. We identify path-graph-like components by utilizing higher-order distribution characteristics of hitting time measures of random graph models. Skewness of the hitting time distribution differentiates synthetic lollipop graphs with path-graph components from fully connected, random, small-world, and scale-free synthetic graphs. We utilize hitting-time distribution to detect sensory pathways during rest and task. In human functional magnetic resonance imaging (fMRI) data, skewness of the hitting time distribution is greater during task than resting state data. Our findings suggest that hitting time skewness can be used to characterize brain-network configuration using path-like features in different task settings and could be diagnostic of some brain disorders.
Industry Application Area(s)
Network Science, Brain network