Artificial Intelligence Boosts Chances of Successful IVF, Study Claims

A new artificial intelligence-enhanced machine can reportedly help doctors select healthy embryos for in vitro fertilization (IVF) treatment with 85 percent accuracy.

The tech—developed by experts from Imperial College in London and Cornell University in New York—uses algorithms to predict the quality of embryos in an incubator based on time-lapse photography. Medical staff can use it to help boost the chance of a successful pregnancy. 

According to The Mirror, a New York–based supercomputer codenamed The Beast was fed “more than 50,000 images of more than 10,000 embryos” to learn the key factors behind which embryos were likely to result in a live birth, and which were likely to result in a stillbirth.

In one trial, using a sample of 328 images of embryos, the AI machine was able to identify 280 that had resulted in a live birth, The Daily Telegraph reported. Nikica Zaninovic, an associate professor of embryology in clinical obstetrics and gynecology and lead author of the study from Cornell University, said the rate was well above what humans could manage.

Zaninovic told The Mirror: “If artificial intelligence [AI] can recognize embryos that are normal and those that are abnormal chromosomally, that will result in a reduced miscarriage and stillbirth rate, which is really our central reason for doing this research. This research is something genuinely new and an 85 percent success rate is huge.”

He said the use of such AI technology may be routine in clinical settings within the next five years.

The embryology expert said using the AI tool could boost the IVF-time-lapse success rate for women under 35 with no underlying health conditions from 50 percent to 70 percent.

Allan Pacey, a fertility expert who works at Sheffield University, told The Telegraph that time-lapse photography had been used in clinics for years—but predictive data showed promise.

“The application of artificial intelligence to the data collected from time-lapse systems is a very good idea as it may find patterns and algorithms that are not visible to the human eye,” he said.

The new technology was presented today during a fertility conference in Denver, Colorado, with the American Society for Reproductive Medicine.

“Our data-driven approach provides a novel way to assess embryo quality and uncovers a new strategy for identifying embryos, which is likely to increase the chance of pregnancy,” the university researchers concluded in a press release shared with Newsweek on Wednesday.

Join the Discussion