The differentiation of multiscale structural gradients from children to adolescents
Poster No:
1235
Submission Type:
Abstract Submission
Authors:
Yirong He1, Qiongling Li1, Debin Zeng2, Lei Chu3, Xiaoxi Dong1, Shuyu Li4
Institutions:
1Beijing Normal University, Beijing, Beijing, 2Beihang university, Beijing, Beijing, 3Beihang University, BEIJING, Beijing, 4Beijing Normal University, Beijing, China
First Author:
Co-Author(s):
Introduction:
The brain development of children and adolescents is accompanied by maturation of cortical myelination and white matter network. Many studies have revealed that cortical maturation follows a pattern of spatiotemporal hierarchy(Sydnor, Larsen et al. 2021). Connectome gradient techniques representing brain topological organization in a low-dimensional space can smoothly capture the cortical macroscale hierarchy(Margulies, Ghosh et al. 2016). A recently introduced in vivo model integrated three features of structural connectivity, including diffusion MRI tractography, geodesic distance, and microstructural similarity(Paquola, Seidlitz et al. 2020). Here, we leveraged this approach to investigate structural connectome development from childhood to adolescence and related it with maturation of cortical morphology and function.
Methods:
Dataset. We collected a longitudinal cohort of 276 participants (aged 6-14 years, 135 females) with 437 scans from Children School Functions and Brain Development Project in China (Beijing Cohort). We obtained structural MRI and diffusion MRI using a 3T Siemens Prisma scanner at Peking University.
Compute multiscale structural connectome gradients. Base on T1w, T2w, and diffusion MRI, we computed geodesic distances, microstructural profile covariance (MPC), and tact strength between brain regions. After connecting the MPC, geodesic matrix, and structural connectivity network horizontally, we calculated the normalized angle similarity between each of the two rows(Paquola, Seidlitz et al. 2020). To compute connectome gradients, the normalized angle matrix was then fed into the diffusion map embedding algorithm, which mapped the high-dimensional multiscale structural connectome data into a low-dimensional space(Coifman, Lafon et al. 2005). We calculated several global features to measure the gradient transitions during development, including gradient range, explanation ratio, standard deviation, and dispersion. By leveraging a mixed effect linear model, we identified the developmental trajectories at both the global level and node-wise level. Meanwhile, we calculated eccentricity measure as Euclidean distance between each node and the centroid of template space.
Association with development of morphometric features. We utilized 5 cortical morphometric measures to investigate the relationships between multiscale structural gradients and morphometric features. We performed principle component analysis (PCA) to summarize these 5 features and related the PC 1 to the multiscale structural gradient 1.
Multiscale structure-function coupling. We conducted an analysis on the coupling between structure and function. Coupling was computed as Spearman rank correlation between connectivity profiles of structure and function.
Compute multiscale structural connectome gradients. Base on T1w, T2w, and diffusion MRI, we computed geodesic distances, microstructural profile covariance (MPC), and tact strength between brain regions. After connecting the MPC, geodesic matrix, and structural connectivity network horizontally, we calculated the normalized angle similarity between each of the two rows(Paquola, Seidlitz et al. 2020). To compute connectome gradients, the normalized angle matrix was then fed into the diffusion map embedding algorithm, which mapped the high-dimensional multiscale structural connectome data into a low-dimensional space(Coifman, Lafon et al. 2005). We calculated several global features to measure the gradient transitions during development, including gradient range, explanation ratio, standard deviation, and dispersion. By leveraging a mixed effect linear model, we identified the developmental trajectories at both the global level and node-wise level. Meanwhile, we calculated eccentricity measure as Euclidean distance between each node and the centroid of template space.
Association with development of morphometric features. We utilized 5 cortical morphometric measures to investigate the relationships between multiscale structural gradients and morphometric features. We performed principle component analysis (PCA) to summarize these 5 features and related the PC 1 to the multiscale structural gradient 1.
Multiscale structure-function coupling. We conducted an analysis on the coupling between structure and function. Coupling was computed as Spearman rank correlation between connectivity profiles of structure and function.
Results:
During development, the first gradient showing differentiation between transmodal and primary regions, and the second gradient separating anterior and posterior regions (Fig 1A). According to the trajectories of global measures, the first gradient increased with development and vice versa for the second gradient. Age-related changes in multiscale structural gradient 1 during development revealed the gradual maturation of the S-A axis (Fig. 1). The first principal component of morphological features was associated with the first gradient, and the age-related patterns of change were also correlated (Fig. 2A-D). The pattern of coupling between multiscale structure and functional connectivity adhered to the S-A axis, and the majority of systems exhibited an increase in coupling with development (Fig. 2E-F).
·Fig. 1 Age-related changes of gradients at both the global level and the node level.
·Fig. 2 Relation to cortical macrostructure and function.
Conclusions:
In conclusion, by applying connectome gradient analysis, we revealed a progressive transition of multiscale structural connectome from childhood to adolescence. Our findings suggested that the organization of structural connectome moves toward a more distributed direction with development, which correlates with maturation of cortical macrostructure and function.
Education, History and Social Aspects of Brain Imaging:
Education, History and Social Aspects of Brain Imaging
Lifespan Development:
Early life, Adolescence, Aging 1
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 2
Novel Imaging Acquisition Methods:
Anatomical MRI
Diffusion MRI
Keywords:
Other - structural connectome; gradients; development
1|2Indicates the priority used for review
Provide references using author date format
Margulies, D. S., S. S. Ghosh, A. Goulas, M. Falkiewicz, J. M. Huntenburg, G. Langs, G. Bezgin, S. B. Eickhoff, F. X. Castellanos, M. Petrides and others (2016). "Situating the default-mode network along a principal gradient of macroscale cortical organization." Proceedings of the National Academy of Sciences 113(44): 12574-12579.
Paquola, C., J. Seidlitz, O. Benkarim, J. Royer, P. Klimes, R. A. I. Bethlehem, S. Larivière, R. V. de Wael, R. RodrÃguez-Cruces, J. A. Hall, B. Frauscher, J. Smallwood and B. C. Bernhardt (2020). "A multi-scale cortical wiring space links cellular architecture and functional dynamics in the human brain." PLoS Biology 18(11).
Sydnor, V. J., B. Larsen, D. S. Bassett, A. Alexander-Bloch, D. A. Fair, C. Liston, A. P. Mackey, M. P. Milham, A. Pines, D. R. Roalf and others (2021). "Neurodevelopment of the association cortices: Patterns, mechanisms, and implications for psychopathology." Neuron 109(18): 2820-2846.