Detecting dynamic changes in modular organization of spontaneous brain activity: A preliminary study

Our brain is a dynamic modular network. Even at rest, brain networks dynamically reconfigure in an organized manner, establishing patterns of connectivity known as resting state networks (RSNs). Recently, significant efforts have been devoted to characterize the dynamics of RSNs. However, little is known about how the dynamic changes in the modular structure shapes the fast spontaneous activity. In this paper, our objective is to validate the feasibility of a recently proposed modularity-based algorithm in investigating RSNs and their variations over time. For this aim, EEG data were recorded from two subjects during resting state. Using EEG source connectivity method with a sliding window, we reconstructed the dynamic brain networks in alpha band. Then, we applied the modularity algorithm to identify the main modular brain states fluctuating over time. The dominant modules were associated with the RSNs. Results showed that the extracted modules were concordant with RSNs found in literature. In particular, the default mode network, known as the most consistent RSN, dynamically alternates its reconfiguration between three modular organizations. Overall, we speculate that this approach, when applied on a larger dataset, will give new insights about the dynamic behavior of RSNs.