Digital Embryology

Infertility affects 25% of couples in NZ and many turn to In Vitro Fertilisation (IVF) to improve their chances of a successful pregnancy. Selection of embryos for implantation is non-trivial due to the complexities of pre-implantation development. Adding to this complexity is the fact that current best practice including subjectively identifying morphological characteristics and developmental timings.

This project aims to: 1) Creating/improving deep learning algorithms that will objectively and reliably annotate embryo images for downstream epidemiological and statistical analyses. 2) Creating ML/AI models to predict clinical pregnancy, fetal heartbeat, and/or live birth 3) Applying ML/AI model to improve our understanding of human embryo development. To achieve these aims you will use the latest in AI/ML techniques such as YOLOv4, Custom DNN, SE blocks, RNN, Random Forests, Boosting, Bagging amongst others.