Longitudinal Development of Brain Functional Connectivity in Early Childhood: Normative Trajectories, Mother-Infant Neural Synchrony, and Brain-Behavior Predictions
Skip to main content
eScholarship
Open Access Publications from the University of California

UCLA

UCLA Electronic Theses and Dissertations bannerUCLA

Longitudinal Development of Brain Functional Connectivity in Early Childhood: Normative Trajectories, Mother-Infant Neural Synchrony, and Brain-Behavior Predictions

Abstract

The early years of life are a critical period for brain development, during which functional connectivity (FC) networks form and refine, shaping future cognitive, emotional, and behavioral capacities. Recent advancements in neuroimaging, including connectome gradients, graph-theoretic approaches, and statistical longitudinal modeling techniques, have enabled a more detailed understanding of large-scale brain network maturation. This thesis leverages a large-scale longitudinal resting-state functional MRI dataset to examine the development of brain FC across the first decade, exploring normative trajectories, at-risk alterations, and predictive markers of later outcomes, while also investigating mother-infant neural synchrony and exploring deep learning models for brain-behavior predictions. The study begins with whole-brain heatmaps quantifying voxel-level FC changes, revealing a nonlinear trajectory, with the most rapid changes in the first year, followed by slower transitions and a transformation phase from two to six years. Then, a comprehensive set of normative FC trajectories is established across the first decade, integrating connectome gradients, maturation scales, and graph-theoretic measures, identifying altered trajectories in at-risk groups and long-term brain-behavior associations. Next, mother-infant neural synchrony is explored, highlighting the ventromedial prefrontal cortex (vmPFC) as a key region for attachment transmission. Findings suggest that prenatal depressive symptoms disrupt, while positive postnatal parenting enhances, intergenerational neural synchrony, emphasizing the importance of early caregiving environments in shaping socioemotional development. Finally, working memory (WM) development is examined, demonstrating a shift from sensory to higher-order networks as the brain matures. Using 3D convolutional neural networks (CNNs), this study promisingly predicts later behavioral outcomes from early FC patterns, particularly in visual and sensory-motor networks, highlighting the potential of early neuroimaging biomarkers for identifying children at risk for neurodevelopmental challenges. Overall, this thesis provides a normative reference and framework for early brain development, identifies neural risk markers, and highlights the potential for early diagnosis and intervention. Future research will refine deep learning models, incorporate graph neural networks (GNNs), and enhance brain connectome analysis to improve predictive accuracy and early screening strategies, ultimately informing interventions that promote optimal neurodevelopmental trajectories in early childhood.