A hybrid premature neonatal segmentation pipeline for clinical brain imaging acquired without dedicated neonatal coils.
Zachary Hill1, Mengyuan Liu1, Sandra Juul2, and Colin Studholme3
1Bioengineering, University of Washington, Seattle, WA, United States, 2Pediatrics, University of Washington, Seattle, WA, United States, 3Pediatrics, Bioengineering, Radiology, University of Washington, Seattle, WA, United States
Due
to the difference in individual cases, larger multi-site studies of
brain injury after premature birth may be needed, but dedicated
neonatal imaging technology isn't
always available. The use of older scanners with coils not
specifically aimed at imaging neonatal brains introduces a severe
intensity variation across the field of view, which
can cause conventional image analysis pipelines to fail. A
robust hybrid tissue segmentation pipeline was developed and shown to
improve tissue segmentations of four test subjects with manual
segmentations for reference. This enables automated and consistent
analysis to better quantitatively study early human brain
development.
This abstract and the presentation materials are available to members only;
a login is required.
