Use of artificial intelligence (AI) embryo selection based on static images to predict first trimester pregnancy loss

Research question Can an artificial intelligence (AI) embryo selection assistant predict the incidence of first trimester spontaneous abortion (SA) using static images of IVF embryos? Design In a blind, retrospective study, a cohort of 172 blastocysts from IVF cases with single embryo transfer and a positive biochemical pregnancy test was retrospectively ranked by the AI morphometric algorithm ERICA™. Making use of static embryo images from a light microscope, each blastocyst was assigned one of four possible groups (Optimal, Good, Fair, Poor), and linear regression was used to correlate the results with the presence or absence of a normal fetal heart rate (FHR) as an indicator of ongoing pregnancy or SA, respectively. Additional analyses included modelling for recipient age and chromosomal status established by preimplantation genetic testing (PGT-A). Results Embryos grouped as Optimal/Good had a lower incidence of SA (16.1%) compared to embryos classified as Fair/Poor (25%, P=0.005, Odds ratio (OR)=0.46). The incidence of SA in chromosomally normal embryos (determined by PGTA) was 13.3% for Optimal/Good embryos and 20.0% for Fair/Poor embryos, although this difference was not statistically significant (P=0.531). There was a significant association between embryo rank and recipient age (P=0.018) in that the incidence of SA was unexpectedly lower in older recipients (21.3% in ≤35, 17.9% in 36-38, 16.4% in ≥39, P=0.0181, OR=0.354). Overall, these results support a correlation between risk of SA and embryo rank as determined by AI; the classification accuracy was calculated to be 67.4%. Conclusions This preliminary study suggests that AI (ERICA™), which was designed as a ranking system to assist with embryo transfer decisions and ploidy prediction, might also be useful in providing information for couples on the risk of SA. Future work will include larger sample size and karyotyping of miscarried pregnancy tissue.