A weighted-RV method to detect fine-scale functional connectivity during resting state

Abstract

During the resting state, in the absence of external stimuli or goal-directed mental tasks, some functionally related discrete regions of the brain show complex low-frequency fluctuations in the blood oxygenation level dependent signal. Here we developed a novel ROI-based multivariate statistical framework to obtain the fine-grained patterns of functionally specialized brain networks in the resting state. Under this framework, the weighted-RV method is proposed and used to detect the spatial fine-scale patterns of functional connectivity. This approach overcomes several major problems of the traditional resting-state data analysis methods such as Pearson correlation and linear regression analysis. By using simulation and real fMRI experiment, we have found that the weighted-RV method is shown to be more sensitive in detecting the fine-scale based low-frequency connectivity even at a very low functional contrast-to-noise ratio (CNR), and this method can achieve much better performance in mapping the fine-grained patterns of functionally specialized brain networks compared to the traditional methods.