Image Processing Reference
In-Depth Information
would be ideal to have more time points for a longitudinal study, this is not
achievable because the infants are extremely preterm and receiving intensive
medical support. Another issue is introduced by the fact that many subjects in
the study could not undergo both sets of scans due to healthcare complications,
which again limits the number of subjects part of a longitudinal study.
Although qualitatively we obtained a very good correspondence for the intra-
subject and inter-subject prefrontal cortex, it is very dicult to validate this
cortical matching, since the data is quite novel and there is no proper way of
establishing a ground truth. The prefrontal cortex undergoes major changes
during this period of time [
5
]. We mapped the folding of the PFC occurring in
this rapid period of development by investigating the change of the mean and
Gaussian curvatures as well as the bending energy required for the folding. The
uniqueness of the application motivated us to choose well-known measures of
curvature and cast these as a measurement of longitudinal change in the indi-
vidual. This is appropriate for a new application of the technique such as here in
contrast to new and unfamiliar measures of curvature. As stated, we have a par-
ticular interest in the bending energy because it can provide information about
the underlying tissue deformation and can be interpreted directly as a measure-
ment of change in the infant folding pattern. The maps of mean and Gaussian
curvature over the entire cohort provide us with information about gyrification
of normal preterm development, and therefore can be used as baseline for further
subjects and identify these infants that do not follow this trend.
This work will be extended to finding correspondences for the whole brain.
However, the folding pattern of the parietal and occipital cortex is much more
advanced than the frontal and pre-frontal region, thus differences may be more
dicult to quantify. Furthermore, the correspondence between the 30 and 40 week
scans can be correlated with diffusion imaging in order to determine the relation-
ship between the mechanism of the surface cortical folding and underlying changes
in connectivity, particularly the establishment of associative cortico-cortical con-
nections which occurs over this 30-40 week period. Combined measurements of
this type have the potential to investigate the mechanical and cellular processes
leading to gyrogenesis. Our immediate future work will investigate correlations
between the rates of cortical folding and neuropsychological outcomes during
infancy and childhood.
Generating accurate correspondences between the intra-subject prefrontal
cortex at multiple time points enables us to measure the longitudinal changes
that take place in this region in preterm infants. These measures may contribute
to early biomarkers for predicting executive and motor development.
Acknowledgements.
This work was supported by funding from the UK char-
ity, SPARKS, EPSRC (EP/H046410/1, EP/J020990/1,EP/K005278), the MRC
(MR/J01107X/1) and the National Institute for Health Research University College
London Hospitals Biomedical Research Centre (NIHR BRC UCLH/UCL High Impact
Initiative).
Search WWH ::
Custom Search