Detecting the Area of Bovine Cumulus Oocyte Complexes Using Deep Learning and Semantic Segmentation

The cumulus-oocyte complex (COC) is an oocyte surrounded by specialized granulosa cells, called cumulus cells. cells surrounding the ensure healthy and embryo development. maturity of COCs at retrieval may be used as indicator to predict outcome assisted reproductive technology (ART). Segmenting a preliminary step in many image processing pipelines evaluate maturity. However, acquiring well-annotated bright-field microscopy datasets remains time-consuming inaccurate procedure, for most biological domains. Additionally, specialists often partially disagree on their annotations, not only among each other, but also own leading inconsistent outcome. Despite recent advancements deep learning segmentation tools biomedical images, there limited usage them having more accurate automated procedures. In this work, we propose pipeline segment bovine images. results our evaluation show that able with same level quality provided human experts.