Temporal Variability of Brain-Behavior Relationships in Fine-Scale Dynamics of Edge Time Series

Abstract

Most work on functional connectivity (FC) in neuroimaging data prefers longer scan sessions or greater subject count to improve reliability of brain-behavior relationships or predictive models. Here, we investigate whether systematically isolating moments in time can improve brain-behavior relationships and outperform full scan data. We perform optimizations using a temporal filtering strategy to identify time points that improve brain-behavior relationships across 58 different behaviors. We analyzed functional brain networks from resting state fMRI data of 352 healthy subjects from the Human Connectome Project. Templates were created to select time points with similar patterns of brain activity. Optimizations were performed to produce templates for each behavior that maximize brain-behavior relationships from reconstructed functional networks. With 10% of scan data, optimized templates of select behavioral measures achieved greater strength of brain-behavior correlations and greater transfer between groups of subjects than full FC across multiple cross validation splits of the dataset. Therefore, selectively filtering time points may allow for development of more targeted FC analyses and increased understanding of how specific moments in time contribute to behavioral prediction. Significance StatementIndividuals exhibit significant variations in brain functional connectivity, and these individual differences relate to variations in behavioral and cognitive measures. Here we show that the strength and similarity of brain-behavior associations across groups vary over time and that these relations can be improved by selecting time points that maximize brain-behavior correlations. By employing an optimization strategy for 58 distinct behavioral variables we find that different behaviors load onto different moments in time. Our work suggests new strategies for revealing brain signatures of behavior.