A new Ovation Fertility/Life Whisperer collaborative study indicates that artificial intelligence technology can evaluate embryo quality more accurately than standard morphological grading methods. When compared to standard embryo morphological grading by
A recent, international collaborative study by U.S.-based Ovation Fertility™ and Australia-based Life Whisperersuggests that artificial intelligence (AI) may be a valuable tool to accurately detect genetic defects in human embryos. In this study, Life Whisperer’s new AI technology was used to evaluate thousands of 2D static images of embryos created through past IVF cycles at Ovation Fertility IVF laboratories. When compared to standard embryo morphological grading by highly skilled embryologists, the AI technology showed a 50% improvement in the prediction of successful implantation– the critical first stage of a successful pregnancy. This is the first study in the U.S. to successfully apply AI by deep learning through computer vision to embryo selection.
The current gold standard for embryo analysis is preimplantation genetic testing (PGT), in which a few cells are removed from each embryo and sent to a specialized genetics laboratory for next generation sequencing. PGT can accurately detect chromosomal abnormalities, so that only chromosomally normal embryos are chosen for transfer to the patient’s uterus. Some U.S. fertility practices now make PGT part of their standard IVF protocols, but not all patients elect to have the procedure performed for financial or other reasons. When PGT is not employed, embryologists select which embryo to transfer based on grading systems that rely on visual inspection of embryos under a high-powered microscope. These study results indicate that AI technology may be an effective tool to help embryologists identify key indicators of embryo morphology which may not be detectable by the human eye.
Using the Life Whisperer AI tool, clinicians drag and drop embryo images taken from a standard microscope camera and get an instant report of embryo viability, helping them determine which embryos are good candidates for PGT or are most likely to result in a successful pregnancy without PGT. For this study, Ovation Fertility provided several thousand photos of embryos from completed IVF cycles, taken before biopsy and transfer. Results of the AI analysis of these images were compared to Ovation Fertility’s documented corresponding chromosomal test result or pregnancy outcome for each embryo, along with the morphological grade assigned by the embryologist.
“The only way to definitively determine if an embryo has genetic defects is PGT, but it is a costly and invasive method,” says Tex VerMilyea, Ph.D., HCLD, Ovation Fertility’s Austin laboratory and operations director. “AI can be a useful pre-screening tool that would allow us to identify and genetically test only those embryos which are viable, with a low likelihood of genetic defects. This can result in significant cost savings for Ovation patients, and also has the potential to improve pregnancy success.”