Meeting Banner
Abstract #2111

Prediction of Vascular Dementia Brain in Distinct Frequency Bandwidths with Whole-Brain Functional Connectivity Patterns

Delong Zhang1, Bo Liu2, 3, Jun Chen2, 3, Xiaoling Peng1, Xian Liu2, 3, Ruiwang Huang1

1Center for Studies of Psychological Application, Guangdong Key Laboratory of Mental Health and Cognitive Science, South China Normal University, Guangzhou, China; 2Department of Radiology, Guangdong Province Hospital of Traditional Chinese Medicine, Guangzhou, China; 3Department of Radiology, Guangzhou University of Traditional Chinese Medicine, Guangzhou, China

Here we used support vector machines (SVM) method to predict the Vascular Dementia (VaD), one of the most common types of dementia, from the functional brain scans according to the whole-brain functional networks, and compared the detection efficiency at different frequency bandwidths, slow-5 (0.01~0.27 Hz), slow-4 (0.027~0.073 Hz), and whole band (0.01~0.073 Hz). The result suggested that whole-brain functional connectivity contain adequate information about neurobiological changes in VaD patients, and the detection efficiency related to the slow-5 was more prominent. These contribute to the understanding of VaD and may facilitate discovery of biomarkers for the diagnosis of individual VaD patient.

Abstract PDF

Keywords

according accuracy adequate allow application atlas attention balance band bandwidths behind brain brains called carried categorical changing characterized chart china classification classifier clinical coefficients cognitive common connectivity constructed contain contains contribute control convergence correlation cross datasets dementia derived detecting detection diagnosis difficulty discovery disease distinct efficiency elements elimination ensemble explored extracted facilitate feature filters finally findings flow forgetfulness frequency functional gets getting global great health healthy hospital incidence include independent indicated individual infarct know laboratory language learning leave linear little machine machines matrices matrix medicine mental motion multivariate networks neuron normalization obviously onset pair participants pass patient patients pattern patterns people predict prediction predictions predictive preprocessing progression prominent psychological radiology realignment recently recruited recursive related removed removing republic scanned scanner science score selection series slice slices slow society south speech spurious steps stepwise studies subject sudden suggest support took traditional trends turned types typically understanding validation variances vascular vector white whole