Experts anticipate widespread use of artificial intelligence in IVF

September 30, 2022

6 minutes to read

Disclosures:
Hariton reports that he serves on the scientific advisory board of Alife and as managing director of the US Fertility Innovation Fund. Monseur does not report any relevant financial disclosures. Please see the literature on all relevant financial disclosures for the authors.


We were unable to process your request. Please try again later. If this issue still occurs, please contact customerservice@slackinc.com.

For several years now, artificial intelligence has been gaining momentum in the clinical environment in specialties such as Kidney diseaseAnd the cardiologyAnd the ophthalmology And the Primary health care.

according to Eduardo Hariton, MD, MBA, A reproductive endocrinologist at the Bay Area Reproductive Science Center in San Ramon, California, the use of artificial intelligence (AI) in reproductive medicine is “currently still in the telephonic phase.” Hariton, who is also vice president of strategic initiatives at US Fertility, estimated that there is a “small minority of early adopters” in IVF who are currently using AI for clinical reasons.


However, it “is likely to play an increasing role in predicting clinical outcomes for patients in individual clinics,” Brent Monsor, MD, SCM, A postdoctoral medical fellow who studies reproductive endocrinology and infertility at Stanford Medicine, Healio.

The benefits of artificial intelligence in artificial insemination

In the field of reproductive health care, the use of artificial intelligence has primarily focused on improving IVF treatment in the form of predicting the potential of eggs, assessing the quality of sperm and eggs, and determining the viability of embryos, Renji Wang and colleagues in a review published in reproduction in 2019.

According to the authors, the selection of embryos is “the single most important factor for IVF success.” However, there is no single definitive criterion that can predict the success of a fetus. Instead, the selection of embryos depends on a variety of factors, making it “difficult to predict the probability of a successful pregnancy for each patient and to fully understand the cause of each failure,” they wrote. The authors suggested that AI may be able to support clinicians in bridging the knowledge gap, thus improving the success rate IVF treatment.

For example, recent data published in Reproductive Biomedicine Online He showed that artificial intelligence – specifically the Life Whisperer Viability algorithm by Presagen – may be able to reduce pregnancy time for people undergoing IVF by 12%. In practice, the algorithm does this by “parsing[ing] Still images of day 5 embryos at the blastocyst stage during IVF procedures,” according to Sonya M. Diacio, Ph.D., The study’s lead author and chief medical sciences officer at Presagen and colleagues.

Whisperer Life viability scores correlate with known traits of fetal quality identified by standard morphological assessment and genetic assessment using [pre-implantation genetic testing for aneuploidy]Diakio said in a press release. However, when comparing the AI ​​with standard morphological scores, the AI ​​performs better in assessing the viability of the fetus. As a result, the use of AI can reduce pregnancy time for couples undergoing IVF treatment.”

In a 2021 opinion piece, Mark B. MD, FACOG, FACS, FacialAnd the He and his colleagues cited reduced gestation time as one of the main benefits of using AI in IVF. Other aspects of the IVF process in which AI can be used to reduce gestation time include “assessment of the quality of gametes, selection of sperm for [intracytoplasmic sperm injection (ICSI)] … recommending patient stimulation protocols, selecting egg donors and alerting the need for maintenance of IVF equipment to name a few,” they wrote.

Similar to Presagen’s embryo selection algorithm, artificial intelligence such as the one used for sperm selection and gamete quality control relies on image analyses.

Leveraging AI in these ways is a more objective approach to IVF than manual selection of embryos, according to Lange and colleagues.

In addition to improving clinical outcomes, AI may relieve some of the monotonous, detailed and time-consuming work of embryologists themselves, so they can “focus on more critical tasks, such as ICSI, embryo biopsy, and training of junior staff,” preventing the “world The damned embryos “burn”, trullis Colleagues wrote.

According to Monseur, “Artificial intelligence can simplify and improve decision-making during IVF cycles as we collect a large number of lab and ultrasound data on a daily basis.”

According to Hariton, AI-enabled simplification may aid the growth of the IVF industry itself.

Eduardo Hariton, MD, MBA

Eduardo Hariton

Other than improving patient outcomes, [AI] very important in our field [because] “We have very severe constraints on the supply side,” Hariton said. We have 1,200 exercises [reproductive endocrinology and infertility specialists] In the United States, we have about 50 graduates each year. The industry’s double-digit growth rate, and the expected size of the market, if everyone had access, would probably be three to four times its size now. And we’re not making decisions to serve this market fast enough.”

The challenges of applying artificial intelligence in practice

In the paper by Wang and colleagues, the authors note that “state-of-the-art [machine-learning] Algorithms like deep learning are still in the preliminary stage and not being adequately researched.” For this reason, most of these machine learning algorithms — a type of artificial intelligence that holds the most promise in IVF, they said — have “ethical and legal risks and issues regarding responsibility, which may lead to [the] distrust of patients and doctors.”

In addition, they cautioned that machine learning models require large amounts of high-quality and unbiased data, or else they may lead to incorrect decisions.

“The data used for predictive models are likely to be clinic-specific and not generalizable to other clinics or a larger number of patients,” Munser noted.

Moreover, “Because a lot of Fertility care individual, and this tool may not work well for all patients.”

In another review published in 2020, Carol Lynn Curshaw, Ph.D., He and colleagues wrote that, “Criteria that appear similar can vary between clinics. For example, if one clinic captures blastocyst images at 110 hours, another may capture them just before freezing—a time that may vary based on fetal growth rate and current workload. in the laboratory “.

Moreover, Trolis and colleagues note that there is no “gold standard” for IVF artificial intelligence.

They wrote: “AI requires calibration, and there is currently no agreement on how to compare the performance of different AI models for optimal methods.”

Curshaw and colleagues also note a lack of prospective research on the implementation of AI in IVF. Although Diakio and colleagues used some future data in their study of Presagen AI, more is needed.

According to Hariton, another challenge in bringing AI into more practice is the current regulatory process.

“When you try to prove that a drug works, you go through a randomized control trial, and you prove that it works against placebo,” Hariton said. “Artificial intelligence is something that you need to continually improve, you need to continually improve. If you go through a two-year clinical trial… you will have an extra two years of data to improve. But you can’t touch the algorithm because if you touch it, you have to go through another clinical trial. something. A randomized control trial is not necessarily the best type of experiment to demonstrate the success of these techniques.”

The future of artificial intelligence in IVF treatment

Much work needs to be done to integrate AI into reproductive medicine and IVF processes in particular. As Curshaw and colleagues note, more research is needed to assess the practical uses of AI in predicting successful IVF pregnancies.

“In a truly futuristic environment, how well AI can select embryos compared to traditional methods is a fundamental question,” they wrote. “Also, is the improvement worth the investment that will inevitably stem from licensing or purchasing AI software and changing and re-validating existing clinical workflows?”

Despite their reluctance, Curshaw and his colleagues see promise in extending AI to IVF due to successes observed in other disciplines—including First Food and Drug Administration (FDA) approval for an AI-based device In April 2021, which is designed to detect colon and rectal polyps.

In fact, the promise of AI has gained so much traction in reproductive medicine that there is now a meeting dedicated to it – the World Fertility Conference with Artificial Intelligence – hosted by Croatia in September.

“It was a meeting with tangible energy,” said Hariton, who attended the conference. Many of these people work for private companies that try to solve similar problems. You can see how that would create this competitive atmosphere. It was just the opposite. Everyone was sharing what they were working on, how they were thinking about problems, and what barriers they were reaching. The collaborative atmosphere at that conference was amazing for a field that was like that [new]. “

As research continues, the AI ​​technology will improve and “may have more uses for determining the quality of eggs, sperm and embryos in conjunction with elapsed imaging,” Monsor predicted.

Hariton agreed, saying he would be surprised if no clinics used AI in some capacity over the next 10 years.

References:

Leave a Comment