Premium
P2‐267: Graph Theoretical Analysis of Functional Connectivity Perturbations in Mild Cognitive Impairment
Author(s) -
Joshi Himanshu,
Bharath Srikala,
Balachandar Rakesh,
Sadanand Shilpa,
Saini Jitender,
Varghese Mathew,
John John P.
Publication year - 2016
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2016.06.1527
Subject(s) - clustering coefficient , cognition , dementia , graph theory , psychology , resting state fmri , graph , correlation , power graph analysis , neuroscience , small world network , cluster analysis , topology (electrical circuits) , complex network , computer science , mathematics , artificial intelligence , medicine , theoretical computer science , combinatorics , geometry , disease
Background: Human brain consists of large, sparse and complex networks characterized by efficient small-world properties which ensures the optimal balance of specialized processing (segregation) and efficient communication (integration) of information. Mild Cognitive Impairment (MCI) is a transition state between normal ageing and Alzheimer’s dementia (AD) where decline in various cognitive domains such as memory, attention, concentration, executive functions and calculation sets in. Examination of functional connectivity at rest during the pre-clinical stage (i.e. MCI) of dementia could throw light on how functional connectivity disturbances could underlie cognitive dysfunction in AD. Methods: Graph theory-based analysis was used to isolate the topological properties of the brain functional networks on the basis of a 200 X 200 graph, G(V,E) where G is a non-zero subset with vertices V1⁄4 fMRI signals from the brain regions and edges E1⁄4 intermodal correlation coefficients as a measure of functional connectivity between nodes. We report normalized clustering coefficients (g), normalized characteristic short path length (l) and smallworldness (s) in 44 MCI and 44 elderly cognitively healthy comparison (eCHC) subjects over network sparsity thresholds of 6 to 30% with an increment of 1% to estimate correlation matrices between 200 brain regions. These network connectivity measures were analysed using brain connectivity toolbox. Results:MCI group in comparison to eCHC group, revealed two main findings: (1) at overall topological level, altered small-worldness and (2) at a nodal topological level, altered nodal characteristics in certain cerebral and cerebellar structures. Using the normalized clustering coefficient (g) values, the brain regions that showed significant differences (p <0.05 FWE-corrected) in MCI group were the right middle frontal gyrus, left superior frontal gyrus, right superior temporal gyrus, right culmen and left tuber of vermis. Conclusions: Efficient small-world