Fig 1.
Post-acquisition processing for volumetric analysis of fetal MRI.
Raw fetal MR images (A) were processed with motion correction and super-resolution volume reconstruction. The rendered images were then aligned in the same direction along the anterior and posterior commissures (B). In coronal and axial views of the reconstructed volume images (C), regional structures of the brain (i.e., the cortical plate, subcortical parenchyma, cerebellar hemispheres, vermis, brainstem, and lateral, third, and fourth ventricles) were manually segmented on each slice (D).
Table 1.
Demographics of fetuses with Dandy-Walker malformation.
Table 2.
Demographics of control fetuses.
Fig 2.
Regional growth trajectories in fetuses with DWM (n = 12) and control fetuses (n = 14).
Growth patterns of each regional volume were modeled and compared between fetuses with DWM and control fetuses. A logarithmic transformation of the volume data was performed before fitting the linear regression model in both the DWM and control groups. The growth trajectory of the vermis (A) was significantly smaller in fetuses with DWM than in control fetuses. Fetuses with DWM exhibited reduced rate of growth in the cerebellar hemispheres (B) and whole cerebellum (C) compared to controls. In contrast, growth trajectories of the left cortical plate (E) and whole cortical plate (F) were significantly larger in fetuses with DWM than in control fetuses. No significant differences were observed between fetuses with DWM and control fetuses in growth patterns of the right cortical plate (D), whole subcortical parenchyma (G), whole cerebrum (H), and whole brain (I).
Fig 3.
Regional growth trajectories in fetuses with DWM (n = 8) and control fetuses (n = 12) in subgroup analysis that excluded fetuses with agenesis of the corpus callosum (ACC).
A logarithmic transformation of the volume data was performed before fitting the linear regression model in both the DWM and control groups. Similar to results of the whole-group analysis (Fig 2), the growth trajectory of the whole cortical plate (F) was significantly larger in fetuses with DWM than in control fetuses. In the subgroup analysis, both the right (D) and left (E) cortical plate were significantly larger in fetuses with DWM than in control fetuses. Further, the growth trajectories of the cerebellar hemispheres (B), subcortical parenchyma (G), whole cerebrum (H), and whole brain (I) were significantly larger in fetuses with DWM than in control fetuses. No significant differences were observed in growth trajectories of the whole cerebellum (C) between fetuses with DWM and control fetuses in the subgroup analysis.
Table 3.
Distribution of the number of vermian lobules in the DWM (n = 7) and control groups (n = 8).