Review of Community Detection in Complex Brain Networks

The brain network community detection algorithm has become a highly regarded topic in recent years within the fields of neuroscience and network science, widely employed to unveil patterns of structural and functional connectivity in the brain. Due to the complexity of the brain networks and the need to handle multiple subjects and various task scenarios, it significantly increases the difficulty of community detection in this field. This paper focuses on functional magnetic resonance imaging (fMRI) technology and comprehensively reviews the advancements in research regarding algorithms for detecting communities within brain functional networks. Firstly, the basic process, task categories, and method types of brain network community detection algorithms are described. Next, various brain network community detection algorithms are classified in different task scenarios, including separate communities, overlapping communities, hierarchical communities, and dynamic community detection algorithms. A detailed analysis of the advantages and disadvantages of different methods is provided, along with their applicable scopes. Finally, the future directions of brain network community detection algorithms are discussed, including the problem of community detection in multi-subject networks, robustness issues in brain network community detection, and studies on brain network community detection algorithms for multimodal imaging data. This paper can serve as a methodological guide for future research on brain network community structures.