Constructing structural and functional brain networks from regional brain profiles

Abstract Brain network analysis has become an important approach to understanding brain function. Given the human brain’s complexity and dynamic nature, traditional brain networks based on temporal synchronization may not capture all the nuances of brain activity. The emerging trend is to construct brain networks from regional information, creating pathways that connect regional profiles to network-level insights to better understand brain function. In this study, we constructed structural networks based on regional brain structural information, specifically gray matter volume (GMV), and functional networks based on regional brain activity, measured by brain entropy (BEN). We compared these newly constructed networks with traditional networks: functional connectivity networks (FCN) based on temporal synchronization, structural connectivity networks (SCN) derived from fiber tracking, and MEG-derived functional networks. Our results reveal that GMV network (GMVN) and BEN network (BENN) show correlations with these networks but also exhibit distinct differences. Furthermore, we conducted connectome gradient analyses, uncovering meaningful brain function distribution patterns in both GMVN and BENN. Finally, we used GMVN, BENN, and traditional FCN to predict cognitive and emotional scores. The results showed that BENN, in the resting state, provided the best prediction of both cognitive and emotional scores. This study systematically evaluates the relationship between brain networks constructed from regional brain information and existing networks, as well as their ability to predict behavioral phenotypes. It demonstrates that networks built from regional information capture aspects of brain activity that traditional networks cannot represent and may even provide superior predictive power for behavioral phenotypes. This opens a new path for understanding brain function from regional to network-level information and lays the foundation for future applications in brain development, individual differences, and clinical research. Competing Interest Statement The authors have declared no competing interest.