Comparing Nonlinear Trajectories Across Brain Networks: A Key to Understanding Complex Brain Dynamics

Abstract

This study explores the nonlinear developmental trajectories of brain networks in neurotypically developing infants during their first six months. Using a longitudinal dataset of 137 resting-state functional MRI scans from 74 infants, we analyzed five spatial metrics across 13 intrinsic connectivity networks, including motor, visual, subcortical, and prefrontal networks. A cubic model was specifically employed to capture distinct linear and non-linear trends in the networks developmental patterns, allowing for the identification of significant differences in cubic, quadratic, and linear slope parameters across networks. This model choice was driven by the need to examine how specific non-linear components (e.g., inflection points and acceleration rates) uniquely characterize each networks trajectory, which a generalized approach might smooth out without pinpointing such network-specific features. Notably, the subcortical network exhibited a distinct cubic growth pattern, while secondary motor and visual networks showed pronounced quadratic variations, suggesting network-specific shifts in spatial organization and connectivity. These findings highlight the unique maturation timelines and interactions between functional systems, such as early sensory-motor coordination and later cognitive integration. The results underscore the importance of network-specific growth patterns, providing deeper insights into how infant brain networks evolve and interact to support emerging cognitive and behavioral functions.