Non-invasive, image-based genetics AI improves live birth outcomes, even for embryos classified as euploid by PGT-A

AI Healthcare company Presagen, in collaboration with US Fertility and Ovation Fertility (USA), presented a study showing that non-invasive, image-based AI analysis of embryo genetic integrity can improve live births outcomes for IVF patients. The genetics AI utilizes static images from a variety of imaging systems as input (both timelapse and non-timelapse) to generate an instantaneous AI score related to the likelihood of the embryo being genetically normal (euploid), which can be used to aid in embryo evaluation on day of transfer or cryopreservation. Importantly, the study showed that ranking embryos using the AI score improved live birth outcomes even when embryos had already been classified as euploid by the invasive preimplantation genetic testing for aneuploidies (PGT-A) method.

In the USA, testing embryos using PGT-A is common practice, as transfer of euploid embryos is thought to be associated with improved pregnancy and live birth outcomes. One of the study’s aims was to assess whether ranking embryos for transfer using the Life Whisperer Genetics AI, which evaluates genetic integrity based solely on AI analysis of embryo morphology, provides any additional clinical benefit for patients when embryos were also classified as euploid according to PGT-A testing. The results showed that regardless of whether embryos had undergone invasive PGT-A, ranking embryos using Life Whisperer Genetics AI scores improved live birth rate on first embryo transfer by 10% for embryos where PGT-A was not performed; and by 7% for embryos classified as euploid by PGT-A. Furthermore, the time to live birth was reduced by 14% for PGT-A tested embryos.

This study suggests that the genetics AI may be identifying additional morphological characteristics associated with genetic integrity beyond the ability of PGT-A, indicating the need to consider both DNA sequence-based and morphology-based genetic evaluation together. The improvement in live birth outcome for PGT-A tested embryos could potentially be due to the AI’s assessment of morphology at the whole embryo level, in contrast to PGT-A which only tests a small sample of biopsied cells that may not be representative of the whole embryo. This hypothesis is the subject of additional research regarding image-based AI analysis of embryo genetic integrity. 

The poster study was presented at the Pacific Coast Reproductive Society (PCRS) 2024 conference in Palm Springs by Dr Matthew “Tex” VerMilyea from US Fertility and Ovation Fertility.