Understanding the rules of life

Bioscience for sustainable agriculture and food

Bioscience for renewable resources and clean growth

Bioscience for an integrated understanding of health

Category: CASE Studentships

CARTAI: Cattle Assisted Reproduction (ART) Facilitated by Artificial Intelligence (AI)

Project No. 2325


Primary Supervisor

Prof Darren Griffin – University of Kent


Dr Peter Ellis – University of Kent

Alejandro Chavez-Badiola – IVF2.0 (CASE Partner)


Cattle Assisted Reproductive Technology (ART) is essential to transport genetics (sperm and/or embryos) worldwide, accelerate genetic gain and biobank rare breeds to preserve biodiversity.

Two billion and one million cattle sperm/embryos respectively are used each year globally, and cattle IVF is also an excellent model for humans.

We seek a student with lab and computer experience to investigate Artificial Intelligence (AI) in cattle ART/IVF, understanding consequences of chromosome abnormalities and DNA damage in gametes/embryos, with an ultimate view to non-invasive genetic screening.

IVF2.0 is a UK-registered company dedicated to improving ART success through AI. Four products (“SiD”, “Air-O”, “SofY”, “ERICA”) select the best sperm (SiD), egg (Air-O), sperm-egg interaction (SofY) and/or embryo (ERICA) most likely to lead to successful IVF pregnancy. AI algorithms are trained against a “ground truth” of successful/unsuccessful outcomes or defined genetic parameters. To date however this has been performed entirely on human material.

Gametes/embryo chromosome abnormality is the leading cause of ART/IVF failure. It is however currently not possible to ascertain whether a sperm is abnormal without destroying the cell. It is possible in eggs/embryos, however the process is controversial as sampling involves removal of cells, which is considered invasive. Sperm DNA damage sperm is one of the most significant and widely ART researched areas, however, we do not know whether we can, by morphokinetics/AI alone select normal sperm. This also impacts embryo quality in that it has to repair damage on the incoming sperm.

We will develop a novel means of selecting both chromosomally normal/abnormal and damaged/undamaged cattle sperm by SiD (and SofY) alone. We have sperm samples with known chromosome rearrangements/DNA damage levels through an active screening programme. Using these sperm to generate abnormal embryos, we will train ERICA and Air-O to differentiate normal/abnormal embryos/eggs, correlating morphokinetic/AI evaluations with known developmental and genetic parameters.