Prefrontal Cortical Folding of the Preterm Brain: A Longitudinal Analysis of Preterm-Born Neonates - Longitudinal Registration and Shape Modeling - page 7

It can be observed however that the Gaussian curvature is very sensitive to
the geometric errors on the surface and thus not as reliable and useful feature
for cortical surface analysis like the mean curvature.
From the maps of change in bending energy (Figure 6 (d)), we can notice
the amount of energy required for the folding gyri and sulci between the two
time points, as anticipated. As mentioned before, the bending energy depends
on both the total curvature difference and the change in local surface area.
The map of local surface area change (Figure 6 (c)) shows that the inferior
prefrontal cortex undergoes more changes than the superior prefrontal cortex
and expands the most overall. The superior cortex undergoes almost no change,
which is expected since it develops and folds mostly before 30 week and only
minor changes happen after [ 5 ].
Fig. 6. Total curvature shown for 31w (a) and 42w (b) EGA, local surface area change
(c) and bending energy required (d) for one subject
Group-Based Analysis at 30 Week GA. A group analysis was performed
by mapping all the the changes into the space of a 30 week EGA and maps of
average change in mean curvature, Gaussian curvature and local area change
were obtained over the cohort.
From the maps of mean and Gaussian curvature (Figure 7 , (a) and (b) ),
we notice the same folding trend as in the individual results, i.e. the same sulci
and gyri are developing, however the overall changes are smaller than in the one
particular case shown above, probably due to the variation of each individual
subject and their gestation age at birth.