Introduction to Artificial Intelligence and History of AI
Automatic segmentation of trophectoderm in microscopic images of human blastocysts
1. AUTOMATIC SEGMENTATION OF TROPHECTODERM IN MICROSCOPIC
IMAGES OF HUMAN BLASTOCYSTS
ABSTRACT
Accurate assessment of embryos viability is an extremelyimportant task in the
optimization of in vitro fertilizationtreatment outcome. One of the common ways of assessing
thequality of a human embryo is grading it on its fifth day of development based
onmorphological quality of its three main components (Trophectoderm, Inner Cell Mass, and
thelevel of expansion orthe thickness of its Zona Pellucida). In this study, we propose afully
automatic method for segmentation and measurement of TEregion of blastcysts (day-5 human
embryos). Here, we eliminatethe inhomogeneities of the blast cysts surface using the
Retimetheory and further apply a level-set algorithm to segment the TEregions. We have tested
our method on a dataset of 85 images andhave been able to achieve a segmentation accuracy of
84.6% forgrade A, 89.0% for grade B, and 91.7% for grade C embryos.