Poster Session E, Monday, March 25, 2:30 – 4:30 pm, Pacific Concourse
Decreased local functional brain connectivity can predict conversion to MCI or dementia
Eun Hyun Seo1, Jinsick Park2; 1Premedical science, College of Medicine, Chosun University, Gwangju, Korea, 2Department of Biomedical Engineering, Hanyang University, Seoul, Korea
Background: It is known that functional brain network is disrupted from the very early stage of Alzheimer’s disease (AD) spectrum. In the current study we investigated functional brain network parameters associated with future cognitive decline in cognitively normal (CN) elderly and individuals with mild cognitive impairment (MCI). Methods: Functional and structural MRI data, clinical and neuropsychological data were downloaded from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. In the final analysis, 52 CN elderly and 48 elderly with MCI were included. Image data preprocessing was carried out using MELODIC of FMRIB’s Software Library. BOLD time course was extracted according to AAL atlas. Based on interregional correlation matrices of 90 ROIs, network parameters were calculated using Brain Connectivity Toolbox. Logistic regression analysis was performed to examine the ability of network parameters to predict conversions. Results: MCI group had reduced network density (p=0.025), clustering coefficients (p=0.036), and global efficiency (p=0.025) but longer path length (p=0.025) than CN group. For followed sample, 15 were converted to dementia or MCI and 58 were nonconverted. Baseline network parameters between converters and nonconverters were significantly different. Logistic regression revealed that age and clustering coefficient significantly predicted the conversion form CN to MCI or MCI to dementia. Discussion and Conclusions: Our findings indicate that resting-state functional network parameters can play important role to predict future cognitive decline. Especially, decreased local functional brain connectivity can predict conversion to MCI or dementia. The current study suggests that inclusion of functional brain network parameters facilitate early detection of clinical progression.
Topic Area: METHODS: Neuroimaging