A Deep Learning Approach to Embryo Quality Assessment
摘要
In this study, we proposed a Deep Learning-based approach to embryo quality assessment at early cleavage and blastocyst stage. We acknowledged that this is the first work conducted utilizing data from a Vietnam hospital with the stated aim in both day 3 and day 5 image data. Our experimental results conclusively show that the proposed approach significantly outperformed traditional classification methods by a margin of at least 9.3% and thereby won the 2023 international ISODS’s Kaggle challenge.