Assessing Graph Properties and Dynamics of the Functional Brain Networks in Alzheimer's Disease

The human brain is the most complex system in nature. It is intrinsically organized into networked system. Theoretical graphic analysis of human brain networks not only sheds new light into the understanding how the human brain works, but also provides information for exploring into neurological and psychiatric disorders. This paper presents our work of MR imaging data and resting-state functional magnetic resonance imaging (R-fMRI) data, characterizing the graph properties related to clustering coefficient, average degree percentage, characteristic path, global efficiency and small-world ness and network dynamics property on the constructed functional brain networks, then, inferring the discrepancies of those properties between patients with Alzheimer's disease and normal controls. Our experimental results demonstrate that functional brain network of normal controls has higher clustering coefficient, average degree percentage and global efficiency and lower characteristic path. Moreover, it also has stronger small-world ness and propensity for synchronization, compared those of functional brain networks of patients with Alzheimer's disease.