Fig. 3: Systems-level cortical maturation in the full imaging sample (n = 199), capturing multi-modal reorganization in association cortices.
A–C Depict analytical approaches. A MT intensities were sampled along 10 equi-volumetric surfaces between gray matter (pial) and gray matter /white matter boundaries to derive microstructural profiles and a microstructural profile covariance (MPC) matrix. B The maturational index (MI) captures correlations between baseline (i.e., a pattern predicted for age 14 by a mixed effects general linear model) and change patterns (i.e., the age effect estimated by that model) in a region’s network. C A low-dimensional axis of MPC age effects was derived by applying diffusion map embedding to order regions according to their similarity in synchronized microstructural differentiation with age. D–G depict derived maturational patterns. D MI of MPC (MIMPC; pFDR <0.05), showing conservative development in ventral temporal and dorsal regions, and disruptive reorganization in fronto-parietal association cortex. E Relationship between the MIMPC and the principal axis of MPC maturation. The non-linear relationship indicates that conservatively developing ventral and dorsal regions follow maximally different developmental patterns. F MI of functional connectivity (MIFC; pFDR <0.05), showing conservative development in sensorimotor cortex and disruptive reorganization in heteromodal association cortex. G Maturational categories: Overlaps between MIMPC and MIFC per Yeo network. + = conservative, - = disruptive. VIS Visual, SM Sensorimotor, DAN Dorsal attention network, VAN Ventral attention network, FPN Frontoparietal network, DMN Default mode network. Gray masks in functional connectivity (FC) data reflect parcels that were excluded due to low signal-to-noise ratios.
