I-Jung Chen1,
Yen-Hsiang Cheng2, Tzu-Cheng Chao1, 2,
Ping-Hong Lai3, 4, Fu-Nien Wang5, Ming-Long
Wu1, 2
1Department
of Computer Science and Information Engineering, National Cheng Kung
University, Tainan, Taiwan; 2Institute of Medical Informatics,
National Cheng Kung University, Tainan, Taiwan; 3Department of
Radiology, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan; 4School
of Medicine, National Yang-Ming University, Taipei, Taiwan; 5Department
of Biomedical Engineering and Environmental Sciences, National Tsing Hua
University, Hsinchu, Taiwan
In general, fMRI analysis includes full spectral width (i.e., bandwidth) without looking into signal changes at different spectral frequency. For a resting-state fMRI (rsfMRI) experiment, low frequency oscillation (0.01 V 0.1Hz) of MRI signals was reported to reveal activities of resting brain networks. In this study, we examined whether effective connectivity among resting brain networks changes at different frequency band. More specifically, conditional Granger Causality (GC) analysis was performed with band pass filtered time series of resting brain networks. The results show that effective connectivity varies in different frequency bands and outflows from each resting-state network present different frequency dependent characteristics.