Development of techniques for Extraction of fetal brain from Magnetic resonance imaging MRI Scans and abnormality detection

Abstract

Segmentation of brain from fetal MRI is a relatively new field, with little work published on fully automatic processing. Automatic brain segmentation methods developed for MRI of newlineadult cannot be directly applied to study the developing fetal brain in utero, since the fetal newlinebrain is significantly different in terms of geometry as well as tissue morphology. In this newlinethesis, we have proposed three segmentation methods, to extract brain portion from the MRI newlineof human fetus and also estimate the volume of the fetal brain at different gestational week. newlineIn our first fetal brain extraction method, we first enhance the contrast of the boundary of newlinethin layers in fetal brain using brightness preserving bi-histogram equalization (BBHE). newlineNext, the region growing process is done by calculating the distance (D) between the center pixel and 8-neighbour pixels of the contrast-enhanced image. newline