Eigenvector Centrality Dynamics From Resting-State fMRI: Gender and Age Differences in Healthy Subjects.

Abstract

With the increasing use of functional brain network properties as markers of brain disorders, efficient visualization and evaluation methods have become essential. Eigenvector centrality mapping (ECM) of functional MRI (fMRI) data enables the representation of per-node graph theoretical measures as brain maps. This paper studies the use of centrality dynamics for measuring group differences in imaging studies. Imaging data were used from a publicly available imaging study, which included resting fMRI data. After warping the images to a standard space and masking cortical regions, ECM were computed in a sliding window. The dual regression method was used to identify dynamic centrality differences inside well-known resting-state networks between gender and age groups. Gender-related differences were found in the medial and lateral visual, motor, default mode, and executive control RSN, where male subjects had more consistent centrality variations within the network. Age-related differences between the youngest and oldest subjects, based on a median split, were found in the medial visual, executive control and left frontoparietal networks, where younger subjects had more consistent centrality variations within the network. Our findings show that centrality dynamics can be used to identify between-group functional brain network centrality differences, and that age and gender distributions studies need to be taken into account in functional imaging studies.