Connectivity in math-gifted adolescents: Comparing structural equation modeling, granger causality, and dynamic causal modeling

Major challenges in brain research include understanding how the brain retrieves, processes, and transmits information along with understanding how information is stored. Therefore, connectivity analyses are vital in exploring information flow and temporal interactions between particular brain regions. This paper presents the results of two different types of connectivity analysis on previously acquired fMRI data of mathematically gifted adolescents and control subjects performing a mental rotation task. It has been hypothesized that mathematically gifted children rely on the parietal region and right hemisphere, along with utilizing inter-hemispheric interactions that may be a more efficient network during mental rotation tasks. Granger causality and dynamic causal modeling (DCM) were used to model the connectivity in the two groups. The model outputs are compared with connectivity paths determined from structural equation modeling (SEM) in a previous study [1]. Although these methods can be used as confirmatory and/or exploratory tools, they may provide complementary, rather than redundant, information about connectivity networks within the brain.