Diffusion tensor imaging-based characterization of brain neurodevelopment in primates

Abstract

Primate neuroimaging provides a critical opportunity for understanding neurodevelopment. Yet the lack of a normative description has limited the direct comparison with changes in humans. This paper presents for the first time a cross-sectional diffusion tensor imaging (DTI) study characterizing primate brain neurodevelopment between 1 and 6 years of age on 25 healthy undisturbed rhesus monkeys (14 male, 11 female). A comprehensive analysis including region-of-interest, voxel-wise, and fiber tract-based approach demonstrated significant changes of DTI properties over time. Changes in fractional anisotropy (FA), mean diffusivity, axial diffusivity (AD), and radial diffusivity (RD) exhibited a heterogeneous pattern across different regions as well as along fiber tracts. Most of these patterns are similar to those from human studies yet a few followed unique patterns. Overall, we observed substantial increase in FA and AD and a decrease in RD for white matter (WM) along with similar yet smaller changes in gray matter (GM). We further observed an overall posterior-to-anterior trend in DTI property changes over time and strong correlations between WM and GM development. These DTI trends provide crucial insights into underlying age-related biological maturation, including myelination, axonal density changes, fiber tract reorganization, and synaptic pruning processes.

Figure 1.

Pipeline of the atlas-based study. (a) A structural atlas was constructed using T₁ images, and different lobar parcellations were defined on this atlas. (b) A corresponding DTI atlas was created with FA images coregistered to T₁ images. (c) ROI-based statistical analysis was performed for each lobar parcellation using a linear growth model. FA curve in the corpus callosum is shown as an example. Tract-based analysis was similar to ROI analysis (c) with the exception that statistical modeling of diffusion properties was done for each sampling point along the fiber instead of in each parcellation ROI. (d) All the fiber tracts included in this study including the genu, the cingulum, the inferior longitudinal fasciculus, different subdivisions of the internal capsule, and the splenium. Fiber tracts were colored by the FA values at postnatal day (PND) 300.

Figure 2.

Estimation of FA, MD, AD, and RD using ROI-based generalized linear model (GLM) at age 300, 900, and 1500 days (ca. 1, 2.5 and 4 years old) for WM and GM, respectively. Visualization was chosen to illustrate the cortical regions of the left hemisphere. Note that color scales were chosen independently for WM and GM to aid visual differentiation of changes over time in different lobar regions. For detailed results for all regions and comparisons between WM and GM, please refer to Figures 3 and 4.

Figure 3.

Estimation of FA, MD, AD, and RD for all the ROIs in WM at age 300, 900, and 1500 days using the linear growth model. Overall changes from 300 to 1500 days are shown as percentages.

Figure 4.

Estimation of FA, MD, AD, and RD for all the ROIs in GM at age 300, 900, and 1500 days using the linear growth model. Overall changes from 300 to 1500 days are shown as percentages. The same scales were used for MD, AD, and RD to aid comparison with WM.

Figure 5.

Correlations of diffusion property changes between WM and GM. Corresponding regions in left and right hemispheres showed highly similar correlation values and P values. Regions that showed a significant P value (<5%) are indicated with *. Note that color scales were chosen to represent the correlation values between the minimum and the maximum and not 0 to 1. Additionally, regions characterized by only WM or GM (e.g., CC, PM, and IL) were shown in black as correlations were impossible to calculate.

Figure 6.

Overlay of results from fiber tract–based and VBA approaches showed consistency of significant findings. Voxels with significant DTI change across time (P values < 0.05) from VBA are highlighted (as red voxels) in the 3D slices while segments of the fiber tracts P values < 0.05 (−lg (P value) > 1.3) from fiber tract analysis are highlighted with the chosen colormap. Arrows point to illustrative regions where significant changes were found consistently with both analyses.

Figure 7.

Significance fiber maps for change across time: FRATS-based P values are mapped along fiber tracts. Colormap was chosen to highlight regions with significant changes over time (P value < 0.05).

Figure 8.

Diffusion properties along the fiber tract were estimated using the developmental data obtained at 300, 900, and 1500 days of age.