Data-driven cortex segmentation in reconstructed fetal MRI by using structural constraints
Abstract
In utero fetal MR images are essential for the diagnosis of abnormal brain development and understanding brain structures maturation. Because of particular properties of these images, such as important partial volume effect and tissue intensity overlaps, few automated segmentation methods have been developed so far compared to the numerous ones existing for the adult brain anatomy. In order to address these issues, we propose a two-step atlas-free cortex segmentation technique including anatomical priors and structural constraints. Experiments performed on a set of 6 in utero cases (gestational age from 25 to 32 weeks) and validations by comparison to manual segmentations illustrate the necessity of such constraints for fetal brain image segmentation.
Origin : Files produced by the author(s)
Loading...