Construction of a neonatal cortical surface atlas using Multimodal Surface Matching in the Developing Human Connectome Project

Abstract

We propose a method for constructing a spatio-temporal cortical surface atlas of neonatal brains aged between 36 and 44 weeks of post-menstrual age (PMA) at the time of scan. The data were acquired as part of the Developing Human Connectome Project (dHCP), and the constructed surface atlases are publicly available. The method is based on a spherical registration approach: Multimodal Surface Matching (MSM), using cortical folding for driving the alignment. Templates have been generated for the anatomical cortical surface and for the cortical feature maps: sulcal depth, curvature, thickness, T1w/T2w myelin maps and cortical regions. To achieve this, cortical surfaces from 270 infants were first projected onto the sphere. Templates were then generated in two stages: first, a reference space was initialised via affine alignment to a group average adult template. Following this, templates were iteratively refined through repeated alignment of individuals to the template space until the variability of the average feature sets converged. Finally, bias towards the adult reference was removed by applying the inverse of the average affine transformations on the template and de-drifting the template. We used temporal adaptive kernel regression to produce age-dependant atlases for 9 weeks (36-44 weeks PMA). The generated templates capture expected patterns of cortical development including an increase in gyrification as well as an increase in thickness and T1w/T2w myelination with increasing age.

Keywords: Cortical surface atlas; MRI; MSM; Neonatal; dHCP.

Figure 1

Histogram of PMA at the time of scan.

Figure 2

Exemplar surfaces and feature sets for an individual scanned at 40 weeks PMA (left hemisphere, medial and lateral views). Left from top to bottom: white matter surface, pial surface, inflated surface, cortical labels; right from top to bottom: sulcal depth map, mean curvature, T1w/T2w myelin maps, cortical thickness (all features shown on the very inflated surface).

Figure 3

Exemplar sulcal depth maps for: an individual projected on very inflated surface scanned at 40 weeks PMA (top left), and Conte69 atlas (top right) projected on very inflated surfaces (left hemisphere). Individual sphere (bottom left) rotates in reference to Conte69 sphere (bottom right) until the sulcal depth features align (bottom middle).

Figure 4

Curvature atlas for week 40 after initialisation (k = 0) and after different number of iterations (k = 1, 5, 10, 20, 30) (top rows). Map of vertex-wise difference of atlas curvature maps between iterations k and k – 1 (middle rows), where Diff represents mean across the brain of absolute vertex-wise difference. (left hemisphere, projected on the very inflated surface). Bottom: plot of mean across the brain of absolute vertex-wise difference.

Figure 5

Final anatomical cortical white matter surface atlas spanning 36 to 44 weeks (left and right hemispheres).

Figure 6

Final anatomical cortical pial surface atlas spanning 36 to 44 weeks (left and right hemispheres, lateral and medial views).

Figure 7

Cortical label atlas (maximum probability version) spanning 36 to 44 weeks on the final template (left hemisphere, lateral and medial views).

Figure 8

Mean curvature atlas spanning 36 to 44 weeks on the final template (data projected on the very inflated surface, left hemisphere, lateral and medial views).

Figure 9

Sulcal depth map atlas spanning 36 to 44 weeks on the final template (data projected on the very inflated surface, left hemisphere, lateral and medial views).

Figure 10

Thickness atlas spanning 36 to 44 weeks on the final template (data projected on the very inflated surface, left hemisphere, lateral and medial views).

Figure 11

T1w/T2w myelin atlas spanning 36 to 44 weeks on the final template (data projected on the very inflated surface, left hemisphere, lateral and medial views).

Figure 12

Weighted average across subjects of absolute sulcal depth values per cortical region (red line - left hemisphere, blue line - right hemisphere; lat.=lateral, med.=medial, ante.=anterior, post.=posterior, inf.=inferior). Error bars represent weighted standard deviation across subjects.

Figure 13

Weighted average across subjects of absolute curvature values per cortical region (red line - left hemisphere, blue line - right hemisphere; lat.=lateral, med.=medial, ante.=anterior, post.=posterior, inf.=inferior). Error bars represent weighted standard deviation across subjects.

Figure 14

Weighted average across subjects of thickness values per cortical region (red line - left hemisphere, blue line - right hemisphere; lat.=lateral, med.=medial, ante.=anterior, post.=posterior, inf.=inferior). Error bars represent weighted standard deviation across subjects.

Figure 15

Weighted average across subjects of T1w/T2w myelin map values per cortical region (red line - left hemisphere, blue line - right hemisphere; lat.=lateral, med.=medial, ante.=anterior, post.=posterior, inf.=inferior). Error bars represent weighted standard deviation across subjects.

Figure 16

Mean of absolute feature map values across the whole brain for individual subjects spanning 36 to 44 weeks (red dots - left hemisphere, blue stars - right hemisphere) and weighted average across subjects of feature map values across the whole brain with error bars representing weighted standard deviation across subjects (red line - left hemisphere, blue line - right hemisphere).

Figure 17

Cumulative sum of percent change of thickness, T1w/T2w myelin maps and surface area between weeks for left (red dots and fitted lines) and right (blue dots and fitted lines) hemispheres.

Figure 18

Weighted average across subjects of surface area values per cortical region (red line - left hemisphere, blue line - right hemisphere; lat.=lateral, med.=medial, ante.=anterior, post.=posterior, inf.=inferior). Error bars represent weighted standard deviation across subjects.

Figure 19

Mean of surface area across the whole brain for individual subjects spanning 36 to 44 weeks (red dots - left hemisphere, blue stars - right hemisphere) and weighted average across subjects of surface area across the whole brain with error bars representing weighted standard deviation across subjects (red line - left hemisphere, blue line - right hemisphere).

Figure 20

Gyrification index (GI) for left and right hemisphere for individual subjects (red dots), with weighted average across subjects and error bars representing weighted standard deviation across subjects (black line), and for the final anatomical cortical pial surface atlas (blue stars and dashed line), spanning 36 to 44 weeks.

Figure 21

Average across subjects of the absolute areal distortion map, spanning 36 to 44 weeks (data projected on the inflated surface, left hemisphere, lateral and medial views).

Figure 22

Average across subjects of the anisotropic deformation map, spanning 36 to 44 weeks (data projected on the inflated surface, left hemisphere, lateral and medial views).