Effective connectivity in the default network using granger causal analysis

Nowadays, there is a lot of interest in assessing functional interactions between key brain regions. In this paper, Granger causality is applied to analyze effective connectivity of the default network in frequency domain. The default network of the brain regions related was constructed by extracting the resting-state time series of functional magnetic resonance imaging (fMRI). Then, selected network nodes were analyzed to compute their significance of causal relationship in the frequency domain. The effective connectivity and node properties of the default network were studied for both stroke patients and normal subjects through in-degree, out-degree and causal density. The experimental results demonstrate that, there are different connectivity characteristics in the default network of stroke patients in different frequency bands, and the effective connectivity is enhanced in some frequency bands compared with that of the normal subjects. In particular, the posterior cingulate gyrus (PCG) exhibits significant connectivity features in the default network. This study proved that the feasibility in using Granger causality analysis to examine effective connectivity within the default network, as well as provide new insights on brain's internal relationships at resting state.