Motion networks predict head motion during rest- and task-fMRI

Abstract

Abstract The capacity to stay still during scanning, which is necessary to avoid motion confounds while imaging, varies markedly between people. Here we investigated the neurobiological underpinnings of head motion using connectome-based predictive modeling and publicly available brain fMRI data from 424 individuals. Two motion networks, comprised of cerebellum and default-mode regions relevant to proprioception (prefrontal and temporal cortices) and interoception (insula), forecasted individual differences in absolute and relative head motion during six different tasks- and two rest-fMRI sessions. These findings, which generalized to a novel group of 1422 individuals, suggest that head motion reflects necessary functional connectivity for exerting inhibitory motor control during scanning. These motion networks could serve as head motion neuromarkers valuable for understanding developmental and disease conditions associated with impaired inhibitory motor control, including restlessness, hyperactivity and behavioral impulsivity.