Can AI Help Improve In-Vitro Fertilization Rates?

Although in vitro fertilization solely involves a sperm and an egg, the process leading to an embryo can be more challenging than just putting the two together.
AI uses colorful, time-lapse heat maps to identify the features of embryos that could result in more successful implants and develop into a pregnancy.

Although in vitro fertilization solely involves a sperm and an egg, the process leading to an embryo can be more challenging than just putting the two together. In 2019, the Centers for Disease Control and Prevention (CDC) reported that approximately 10% of women in the U.S. between the ages of 15-44 had difficulty either getting pregnant or avoiding a miscarriage. This demographic of women suffering from ovulation disorders directly translates into a potential patient population for assisted reproduction procedures.

According to one Israeli machine learning startup, AI has the potential to boost the success rate of in vitro fertilization (IVF) by as much as 3x compared to traditional methods. As the number of couples struggling to conceive continues to grow,  the use of AI could substantially increase the pregnancy rates and drive down the prices of IVF’s and related devices.

Assisted reproduction technologies (ART), such as IVF, help women or couples conceive children by removing eggs from a woman’s ovaries, fertilizing them with sperm, and then implanting it back in the body. Most couples who use ART are affected by infertility, although ART can also be used to help lesbian couples or single women.

Artificial Intelligence Used for In Vitro Fertilization

AiVF is using machine learning (ML) and computer vision technology to allow embryologists to discover which embryos have the most potential for success during intrauterine implantation. AiVF’s ML technique involves creating time-lapse videos of developing embryos in an incubator. Recorded over five days, the video shows the milestones of embryo development. With COVID19 disrupting the course of many gynecological procedures, AI can become a safer and more cost-effective solution for struggling couples in 2021 and onwards.

“By analyzing the video, you could dig out so many milestones and so many features the human eye cannot even detect, basically you train an algorithm on successful embryos, and you teach the algorithm what are successful embryos.”

Daniella Gilboa, CEO of Tel Aviv, Israel-based AiVF

American Gynecological Device Market Size and Analysis

In 2020, the total U.S. Assisted Reproduction Technology Device Market size was valued at over $36 million. This market was dominated by Cooper Surgical holding over 50% market share, followed by Cook Medical and Irvine Scientific.

Fertility is a particularly good market for machine learning because the data is clean, with binary outcomes. Other healthcare uses of AI include analyzing pathology or radiology images, but these deal with “a lot of gray zones.” With IVF, you’re either pregnant, or you’re not pregnant. In addition to this market showing great potential, with the use of AI it has also been showing tremendous growth as the average age of couples seeking to have children has been steadily increasing. The average age of first birth in the U.S. has risen from 21.4 in 1970 to 26 in 2015. As couples get older, their fecundity often decreases, thus increasing the need for the use of assisted reproductive technologies.

Register to receive a free US Market Report Suite for Gynecological Devices 2020 – 2026 synopsis

Currently, one of the biggest market limiters for gynecological devices is the high costs involved with these procedures. The cost of all the devices, as well as other healthcare costs, can increase the cost to more than $10,000 per cycle. Many people who are seeking to reproduce either cannot afford multiple cycles or cannot afford it at all. This will continue to limit the market until the cost of assisted reproduction decreases. However, with the new developments in AI detecting more reliable cycles, this could push the market through that limitation. 
Via: IEEE Spectrum

Leave a Reply