When
Where
Abstract: The brain is comprised of interacting neurons, and its complexity poses significant challenges for researchers to understand its structure, function, and dynamics. Statistical network analysis (SNA), a collaboration between network science and statistics has emerged as a powerful tool to understand the generative process of such interconnected systems and integrate multimodal brain imaging data. In this talk, I present two methodology innovations under SNA to improve the current landscape for imaging biomarker detection. With the latent variables-assisted statistical network analysis (LatentSNA), we substantially improve the statistical power for identifying biomarkers; and as a result, we discover large, star-like brain functional architectures implicated in the development of internalizing symptoms in 5,000 to 7,000 children of the Adolescent Brain Cognitive Development (ABCD) study. This finding supports internalizing as a complex and involving psychological phenomenon whose development involves large-scale affective interference of multiple coordinating functional systems. The proposed methods have broad applicability and can contribute to many domains of science with rigorous and powerful analysis.