AI effective at measuring blastocyst morphometric parameters | Image Credit: © ipopba - © ipopba - stock.adobe.com.
Automatically measured blastocyst morphometrics can be used to improve blastocyst selection, according to a recent study published in Scientific Reports.
- The rate of pregnant women under 35 years treated with in-vitro fertilization (IVF) has risen over the years, with implantation rates from non-donor oocytes increasing from 27.6% to 41.6%.
- A blastocyst grading system has been developed, focusing on blastocyst expansion level and the integrity of trophectoderm and inner cell mass.
- The use of a time-lapse monitoring system with artificial intelligence tools has shown promise in annotating morphokinetic events, identifying blastocyst morphology, and selecting embryos with better quality.
- Researchers conducted a retrospective nested case–control study involving women who received IVF treatment between 2014 and 2017. They focused on day-5 blastocyst transfers monitored using an Embryoscope, excluding certain cases such as frozen embryo transfers or embryos from donor oocytes.
- The study found that automated measurements of blastocyst morphometrics revealed associations between blastocyst expansion and implantation rate, as well as a negative correlation between a woman's age and implantation.
From 2003 to 2023, the rate of pregnant women aged under 35 years treated with in-vitro fertilization (IVF) has increased, with rates of implantation from non-donor oocytes increasing from 27.6% to 41.6%.
Optimizations of embryo culture conditions contributing to this increase include extended embryo culture for up to 6 days. This delays embryo transfer to the blastocyst stage, improving uterine and embryonic synchronicity and increasing the rate of live births.
A blastocyst grading system was developed to help identifyhigh-quality blastocyst by focusing on blastocyst expansion level and trophectoderm (TE) and inner cell mass (ICM) integrity. Blastocyst diameter, width, and area have been positively associated with clinical pregnancy rate, and TE quality with implantation rate and live birth rate.
Studies have found a time-lapse monitoring (TLM) systemin an artificial intelligence (AI) tool can annotate morphokinetic events, detect blastocyst morphology, and identify embryos with greater blastocyst quality. However, debates have persisted on the applicability of TLM.
To evaluate a novel approach in blastocyst analysis using an AI tool, investigators conducted a retrospective nested case–control study. Participants included women receiving IVF treatment from 3 public IVF units from 2014 to 2017.
Eligibility criteria included receiving a day-5 blastocyst transfer IVF procedure monitored using an Embryoscope, having known implantation data, having 1 or more transferred embryos, and transferred embryos resulting in no implantation. Women with frozen embryo transfer cycles, transfers with preimplantation genetic testing, or embryos from donor oocytes were excluded.
The gonadotropin-releasing hormone antagonist and the long gonadotropin-releasing hormone agonist protocols were used for ovarian stimulation. Fertilization was accomplished using insemination or intracytoplasmic sperm injection.
There were 608 day-5 transferred blastocysts included in the final analysis, 32.9% of which had a positive known implantation data (KIDp) and 67.1% a negative KID (KIDn). Patients were aged an average 33.5 years, with KIDp embryos associated with a younger maternal age than KIDn embryos, at 30.9 years and 34.8 years, respectively.
Significantly larger blastocyst sizes were observed among KIDp embryos compared to KIDnembroys. A smaller ICM-to-blastocyst size ratio was seen in embryos leading to implantation compared to those without implantation.
A significant positive correlation was also observed between blastocyst size and implantation, with an increase in blastocyst size of 1 μm associated with a 2.1% relative increase in the odds of implantation. Age was negatively associated with the odds of implantation, and ICM did not significantly differ between implanted and nonimplanted embryos.
The automated measurements of blastocyst morphometrics determined an association between blastocyst expansion degree and implantation rate, as well as a negative correlation between woman age and implantation. Investigators concluded blastocyst selection can be improved using automatically measured blastocyst morphometrics.
Fruchter-Goldmeier Y, Kantor B, Ben-Meir A, et al. An artificial intelligence algorithm for automated blastocyst morphometric parameters demonstrates a positive association with implantation potential. Scientific Reports. 2023. doi:10.1038/s41598-023-40923-x