Synopsis

We have developed an approach to elucidate the macrostructural changes of individual white matter tracts in neonatal brains, using deformation-based morphometry. Many studies have investigated how microstructure changes during development, however it is not clear how to disentangle changes in tissue microstructure from macrostructural volume changes. The latter is particularly challenging in the absence of longitudinal data from the same subject, which is the norm given the difficulty in scanning neonates. We propose an approach that is aimed towards this problem. Results are presented from analysis carried out on publicly available diffusion MRI data from the developing Human Connectome Project.

Introduction

Characterising typical early white matter development is a crucial first step towards understanding the links between structure and function in the developing brain^1,2. Many studies have investigated how tissue microstructure changes during development^1,3,4, however it is unclear how to disentangle changes in MRI-derived microstructure indices from macrostructural volume changes. The latter is particularly challenging in the absence of longitudinal data from the same subject, which is the norm given the difficulty in scanning neonates. Here, we propose an approach that is aimed towards this problem. We define subject-specific tract ROIs generated from standardised tractography procotols. We then employ features from non-linear warp fields to estimate volume changes with respect to a baseline. We use these to capture micro and macro-structural changes within individual white matter tracts.

Methods

We used diffusion MRI data from the first public release of the Developing Human Connectome Project (dHCP). Specifically, 40 neonatal subjects (15 females, ages 37 to 44 weeks post conception), were imaged during natural sleep at 3T with a dedicated neonatal imaging system^5,6. Pre-processing and distortion correction was carried out according to the dHCP pipeline⁷. Tractography was performed using standard space protocols to obtain 25 (11 bilateral) tracts across subjects. Fibre orientations were estimated using a parametric deconvolution approach⁸ available in FSL. Probabilistic tractography was then used to reconstruct WM tracts in each subject using protocols defined in template neonatal space⁹ with pre-specified seeds, exclusion masks, and waypoints^7,10. The obtained tracts are shown in figure 1. A diffusion tensor model was fitted to the data from the b = 400 s/mm² and 1000 s/mm² shells and FA and MD maps were obtained. Average microstructural parameters were calculated over a binary mask of each tract in diffusion space (thresholded path probability of 0.005). Deformation-based morphometry was used to assess tract growth, as illustrated in figure 2. Non-linear warps from each subject's native dMRI space to a standard 37-week template space⁹ were generated with Advanced Normalisation Tools (ANTs)¹¹. We produced the Jacobian, J, of each warp field - a tensor at each voxel that describes the positions of the data in relation to the template. The absolute value of the determinant of the Jacobian (|det(J)|) at each point gives the factor by which the local volume expands (|det(J)|>1) or shrinks (0<|det(J)|<1) due to the transformation. The median 1/|det(J)| within each tract was used as a measure of the relative volume change compared to the template. A generalized linear model was then used to assess the growth of each tract with age, with total brain volume included as a regressor.

Results and Discussion

We first performed a regression between age and tract-specific microstructural indices. The results are shown in figure 3, with the heights of the bars representing the standardised regression coefficients. FA shows a significant increase with age for each tract, while MD shows a significant reduction. There was good correspondence between the rates of development characterised by FA and MD. The most rapidly developing tract was the forceps minor, and the most slowly developing was the cingulum. This is in agreement with results from previous studies, in which the cingulum was found to be among the most slowly developing fibre bundles in terms of both myelination¹² and dMRI-derived FA and MD³. The standardised regression coefficients between age and volume changes are shown for each tract in figure 4. Eleven of the tracts showed significant positive correlation with the average 1/|det(J)|, after correction for multiple comparisons, indicating volumetric growth. An opposite trend was observed in one tract, the middle cerebellar peduncle, although this was not significant after correction for multiple comparisons. This agrees with a similar trend reported in^13,14, but warrants further investigation. For some tracts, such as the inferior fronto-occipital fasciculus and the inferior-longitudinal fasciculus, both microstructure and volume changes happen quite rapidly. For others, volume changes do not necessarily align with microstructural development. For example, we did not observe significant volumetric growth in the corticospinal tract, but it displayed above-average rate of microstructural maturation. These trends are in agreement with the expectation that macroscopic developmental changes are separate from myelination and axonal density increase, as there are several different processes contributing to white matter volume increases, including developing vasculature and proliferation of glial cells¹⁵.

Conclusion

Acknowledgements

References

1. Counsell, S. J. et al. Specific relations between neurodevelopmental abilities and white matter microstructure in children born preterm. Brain 131, 3201–3208 (2008).

2. Hagmann, P. et al. White matter maturation reshapes structural connectivity in the late developing human brain. Proc. Natl. Acad. Sci. U. S. A. 107, (2010).

3. Dubois, J. et al. Asynchrony of the early maturation of white matter bundles in healthy infants: Quantitative landmarks revealed noninvasively by diffusion tensor imaging. Hum. Brain Mapp. 29, 14–27 (2008).

4. Zanin, E. et al. White matter maturation of normal human fetal brain. An in vivo diffusion tensor tractography study. Brain Behav. 1, 95–108 (2011).

5. Hughes, E. J. et al. A dedicated neonatal brain imaging system. Magn. Reson. Med. 78, 794–804 (2017). 6. Hutter, J. et al. Quiet echo planar imaging for functional and diffusion MRI. Magn. Reson. Med. (2017). doi:10.1002/mrm.26810

7. Bastiani, M. et al. Automate pre-processing pipeline and quality control for neonatal diffusion MRI in the developing Human Connectome Project (dHCP). in ISMRM 1269 (2017).

8. Sotiropoulos, S. N. et al. Fusion in diffusion MRI for improved fibre orientation estimation: An application to the 3T and 7T data of the Human Connectome Project. Neuroimage 134, 396–409 (2016).

9. Serag, A. et al. Construction of a consistent high-definition spatio-temporal atlas of the developing brain using adaptive kernel regression. Neuroimage 59, 2255–2265 (2012).

10. de Groot, M. et al. Improving alignment in Tract-based spatial statistics: Evaluation and optimization of image registration. Neuroimage 76, 400–411 (2013).

11. Avants, B. B. et al. A reproducible evaluation of ANTs similarity metric performance in brain image registration. Neuroimage 54, 2033–2044 (2011).

12. Kinney, H. C., Brody, B. A., Kloman, A. S. & Gilles, F. H. Sequence of Central Nervous System Myelination in Human Infancy II. Patterns of Myelination in Autopsied Infants. J. Neuropathol. Exp. Neurol. 47, 217–234 (1988).

13. Choe, M. S. et al. Regional infant brain development: An MRI-based morphometric analysis in 3 to 13 month olds. Cereb. Cortex 23, 2100–2117 (2013).

14. Re, T. J. et al. High-angular resolution diffusion imaging tractography of cerebellar pathways from newborns to young adults. Brain Behav. 7, e00589 (2017).

15. Dubois, J. et al. The early development of brain white matter: A review of imaging studies in fetuses, newborns and infants. Neuroscience 276, 48–71 (2014).