Individualized network topography in pre-adolescent children and adults using naturalistic precision fMRI

Abstract

Group-averaged network definitions, commonly used in developmental functional connectivity research, limit our understanding of how network topography may change with age and can lead to inaccurate estimates of age effects on intra- and inter-network functional connectivity. Here, we collected a precision fMRI dataset from 24 parent-child pairs (children 6 to 8 years, 13 females; adults 33 to 47 years, 12 females) during passive video viewing and derived individual template-matched functional network maps from 60 minutes of motion-censored data per participant. Overall, large-scale network architecture was broadly shared between children and adults, with no age-effects observed for network-level surface area, few age differences in network boundaries and generally lower assignment confidence in children. Considering inter-individual similarity in network topography, association networks showed stronger within-family similarity and no overall effect of age on similarity of networks between pairs of individuals. We further asked how individualized network definitions might impact estimates of age effects on dense functional connectivity. While all approaches pointed to greater within-network functional connectivity in adults, we found that individualized approaches had larger age effects and lower sensitivity to head motion. Together, our results suggest that, relative to adults, pre-adolescent children show reduced network assignment confidence and weaker within-network connectivity, but limited differences in network borders and size, and underscore the value of individualized mapping for increasing sensitivity to age effects.