Deciphering the Dynamic Spatiotemporal Maturation from Childhood to Adolescence

Presented During:

Thursday, June 27, 2024: 11:30 AM - 12:45 PM
Room: ASEM Ballroom 202  

Poster No:


Submission Type:

Abstract Submission 


KYOUNGSEOB BYEON1, Shinwon Park1, Jon Clucas1, Kahini Mehta2, Matthew Cieslak2, Seok-Jun Hong3, Hyunjin Park3, Jonathan Smallwood4, Theodore Satterthwaite2, Michael Milham1, Ting Xu1


1Child Mind Institute, New York, NY, 2University of Pennsylvania, Philadelphia, PA, 3Sungkyunkwan University, Suwon, South Korea, 4Queen’s University, Ontario, Canada

First Author:

Child Mind Institute
New York, NY


Shinwon Park  
Child Mind Institute
New York, NY
Jon Clucas  
Child Mind Institute
New York, NY
Kahini Mehta  
University of Pennsylvania
Philadelphia, PA
Matthew Cieslak  
University of Pennsylvania
Philadelphia, PA
Seok-Jun Hong  
Sungkyunkwan University
Suwon, South Korea
Hyunjin Park  
Sungkyunkwan University
Suwon, South Korea
Jonathan Smallwood  
Queen’s University
Ontario, Canada
Theodore Satterthwaite  
University of Pennsylvania
Philadelphia, PA
Michael Milham  
Child Mind Institute
New York, NY
Ting Xu  
Child Mind Institute
New York, NY


Cortical maturation from childhood to adolescence plays a crucial role in neurodevelopment, shaping cognition, emotions, and behaviors. Convergent evidence suggests that neurodevelopment proceeds in a hierarchical manner, with heterogeneous structural and functional maturation patterns. However, the relationship between the established static functional patterns and the brain's intrinsic spatiotemporal dynamics remains underexplored. To address this gap, we employ Complex Principal Component Analysis (CPCA), a technique capable of reducing the complexity of high-dimensional spatiotemporal data on multiple development datasets. In this study, we aim to understand how spatiotemporal patterns develop with age, both locally and globally, focusing on three distinct propagation pathways from childhood to adolescence.


We utilized resting-state fMRI data across a developmental continuum, including Human Connectome Project Development (HCP-D, ages 8.1-21.9, n=408) and Aging (HCP-A, ages 36.0-64.9, n=399) datasets, NKI-Rockland Sample (n=839), and CCNP dataset (n=152) cohorts. The preprocessed data were parcellated followed by CPCA to extract components (i.e. spatiotemporal patterns), aligning them based on their spatial similarity with static functional gradients. Utilizing a dual-regression framework, we reconstructed individual dynamic patterns and assessed the reproducibility across cohorts, test-retest reliability across sessions, and age effects by comparing the individual patterns with the adult reference from HCP-A dataset. Additionally, we examined the age effect of amplitude for the temporal fluctuations and decoded using NeuroQuery and gene ontology enrichment analysis.
Supporting Image: F1.png


We identified phase-dependent spatiotemporal components and focused on the first three dynamic states (Fig. 1A). These were reproducible across cohorts and each explained over 10% of the variation in each datasets (Fig 1B). High test-retest reliability was observed for the first three patterns (r > 0.9), revealing individual-specific patterns can be differentiated from other participants (Fig. 1C). Compared to the adult reference, a linear increased similarity with age (r=.36-.56) was observed, indicating maturation towards adult-like dynamic states (Fig. 1D). As shown in Fig 2A, occurrence time ratio (OTR), representing the time spent in each dynamic state, exhibited a positive age effect for pattern 1 (r=.25) but a negative age effect for pattern 2 (r=-.14). A strong correlation (r=.93) was observed between OTR and the EVR, suggesting the significant contribution of time spent in each state to their altered EVR. We also calculated the proportion of bottom-up (sensory-to-association) and top-down (association-to-sensory) contributions to the first dynamic pattern. Interestingly, with increasing age, individuals spent less time in bottom-up but more time in states characterized by top-down propagation (Fig. 2B). Additionally, the amplitude of regional fluctuations for each dynamic state varied across individuals. From childhood to adolescence, amplitude increased in specific brain regions for patterns 1 and 3, while it decreased for pattern 2. (Fig. 2C), which appears to be related to cell proliferation, protein synthesis, and metal ion regulations.
Supporting Image: F2.png


Our study characterized reproducible and reliable spatiotemporal dynamic patterns across cohorts, establishing their developmental trajectory from childhood to adolescence. We observed the progressive maturation of dynamic brain states across development, with both sensory-centered and hierarchical propagations present in childhood. However, from childhood to adolescence, hierarchical top-down propagations become more common suggesting prior work examining cortical development using static measures of functional connectivity may instead reflect a developmental trend towards the emergence of states where brain dynamics are maintained via hierarchical propagation.

Lifespan Development:

Early life, Adolescence, Aging 1

Modeling and Analysis Methods:

fMRI Connectivity and Network Modeling 2


Other - Hierarchical organization; Spatiotemporal dynamics

1|2Indicates the priority used for review

Provide references using author date format

Bolt T, ‘A parsimonious description of global functional brain organization in three spatiotemporal patterns’, Nature neuroscience. 2022 Aug;25(8):1093-103.
Ge SX, ‘ShinyGO: a graphical gene-set enrichment tool for animals and plants’, Bioinformatics. 2020 Apr 15;36(8):2628-9.
Markello RD, ‘Standardizing workflows in imaging transcriptomics with the abagen toolbox’, elife. 2021 Nov 16;10:e72129.
Margulies DS, ‘Situating the default-mode network along a principal gradient of macroscale cortical organization’, Proceedings of the National Academy of Sciences. 2016 Nov 1;113(44):12574-9.
Schaefer A, ‘Local-global parcellation of the human cerebral cortex from intrinsic functional connectivity MRI’, Cerebral cortex. 2018 Sep 1;28(9):3095-114.
Sydnor VJ, ‘Neuroimaging of plasticity mechanisms in the human brain: from critical periods to psychiatric conditions’, Neuropsychopharmacology. 2023 Jan 1;48(1):219-20.