Judging the quality of embryos for cryopreservation or transfer is an ‘art’ in cattle ART! With MOET embryos, experienced cattle embryologists can ‘pick out’ the ones that obtain a 60% pregnancy rate or higher on transfer. Grading systems have been developed and assessed for MOET embryos over many decades and the International Embryo Technology Society (IETS) provides an excellent resource to assist grading, based on a simple 3 point quality score (some use a 4 point system).
But what about cattle IVP embryos? Most would argue that pregnancy rates are about 10% or more behind MOET embryos. So should we apply the same scoring to IVP embryos as applied to MOET? Should we be more stringent in grading IVP embryos, if the aim is to maximise pregnancy rate? What aspect of morphological grading of IVP embryos is most associated with successful pregnancy outcome: Colour and density? The level of compaction/blastulation? Relative size, shape and proportion of the ICM cells?
One solution is to follow what is happening in human clinical IVF, which is rapidly adopting machine learning algorithm-based methods for analysing images of embryos for their pregnancy success capacity. However, this required the availability of many thousands of embryos imaged, with their subsequent pregnancy outcome recorded. This is virtually impossible for cattle IVP, as who takes thousands of images to build robust algorithms?
Fortunately, The University of Adelaide researchers joined forces with a specialist machine learning team, and have amazing results for predictive pregnancy on much less numbers of embryos, which was obtained exclusively using ART Lab Solutions media suite. Excitingly, the same team have substantiated the pilot study with a second study.
Contact firstname.lastname@example.org or visit www.artlabsolutions.com for more information.