Microstructural correlates of infant functional development: example of the visual pathways

Abstract

The development of cognitive functions during childhood relies on several neuroanatomical maturation processes. Among these processes is myelination of the white matter pathways, which speeds up electrical conduction. Quantitative indices of such structural processes can be obtained in vivo with diffusion tensor imaging (DTI), but their physiological significance remains uncertain. Here, we investigated the microstructural correlates of early functional development by combining DTI and visual event-related potentials (VEPs) in 15 one- to 4-month-old healthy infants. Interindividual variations of the apparent conduction speed, computed from the latency of the first positive VEP wave (P1), were significantly correlated with the infants' age and DTI indices measured in the optic radiations. This demonstrates that fractional anisotropy and transverse diffusivity are structural markers of functionally efficient myelination. Moreover, these indices computed along the optic radiations showed an early wave of maturation in the anterior region, with the posterior region catching up later in development, which suggests two asynchronous fronts of myelination in both the geniculocortical and corticogeniculate fibers. Thus, in addition to microstructural information, DTI provides noninvasive exquisite information on the functional development of the brain in human infants.

Figure 1.

a, Reconstruction by tractography of the optic radiations, with three segments. b, Visual evoked response to faces: grand-average across all infants measured across six electrodes around Oz, and P1 topographic map at 140 ms. ant, Anterior; mid, middle; post, posterior.

Figure 2.

Changes in conduction speed across the infant group with age and DTI indices in the optic radiations. a–c, For the models detailed in Table 1, with age and DTI index covariates (FA, λ_⊥, λ_//), the plots represent the variations of residual speed, as a function of age after correction for the DTI index effect (left column, speed − βDTI index) and as a function of DTI index after correction for the age effect (right column, speed − αage). The contribution of the DTI index was significant for FA and λ_⊥, but not for λ_//. ns, Not significant.

Figure 3.

Heterogeneous maturation within and along the optic radiations. Left and right columns represent normalized FA and λ_⊥, respectively. a, Group mean indices, with quantification on average over the three segments (with SDs in plot bars). Advanced maturation was observed in the anterior segment, with indices closest to 1, followed by the posterior segment, with lower nλ_⊥ compared with the middle segment. ant, Anterior; mid, middle; post, posterior. b, Age-related changes of indices, showing an increase in nFA in the posterior segment and decreases in nλ_⊥ in all segments. c, Quantification of normalized indices along the tracts, from the LGN (abscissa, 0) to the occipital pole (abscissa, 1): mean over the infant group and results for subjects of different ages (5.6, 9.7, and 17.1 weeks of age). The topographical patterns were relatively similar across infants (normalized indices approach 1 with age and maturation), except at the level of the posterior segment where the interindividual variability was high (arrows), with a strong increase of nFA and decrease of nλ_⊥ with age.