Hi! Please leave us your message or call us at 01.800.123.456
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form
Could A.I. Be the Key to Improving IVF Success Rates?
Advancements in artificial intelligence have simplified and standardized IVF processes that have historically been costly and complicated. No longer needing to rely only on human judgement, embryologists can now increase the accuracy of their work with the assistance of A.I. systems.
February 17, 2021
When it comes to the success of In Vitro Fertilization, the quality of the implanted embryos is extremely important, and the selection of the best embryos is a key part of treatment. Although doctors have traditionally been the judges and evaluators of embryo quality, researchers are finding that A.I. can outperform the accuracy of doctors in the identification and selection of embryos that would be most likely to result in a healthy birth.
This new information leads to one main question: what do advancements in A.I. mean for the future of IVF?
Why Embryo Selection is Important
In order for IVF treatment to result in pregnancy and a birth, the inserted embryo must successfully implant into the uterus and must then develop a strong placenta. When determining which embryos to use for treatment, doctors consider the health and stability of a selection of embryos to ensure the highest chance of implantation and corresponding development.
Because IVF treatment has an approximately 30% success rate, and because treatment can be demanding in both physical and financial ways, identifying the most viable embryo plays a big part in increasing the chances of IVF success and preventing the need for multiple rounds of treatment.
In the past, embryologists have been the ones to identify and pick the most promising embryos. But with the help of new A.I. technology, it is seeming more and more likely that this embryo identification could become even more efficient and accurate.
Artificial Intelligence and IVF
Recent advancements in artificial intelligence specifically suited for IVF processes have been taking place at Brigham and Women's Hospital as well as Massachusetts General Hospital, both of which are in the Boston area. At these medical campuses, researchers and investigators are in the process of developing A.I. systems that will assist embryologists with the objective selection of embryos for use in IVF treatment.
But how do these systems actually work?
Data Collection and Deep Learning
In the context of artificial intelligence, “deep learning” refers to an A.I.’s ability to imitate the mechanisms of the human brain in creating patterns and processing data for the purpose of decision making. If an A.I. has a deep learning function, it is able to detect objects, recognize speech, and then use this information to make decisions. Essentially, a deep-learning A.I. can learn without human supervision and can interpret unstructured or unlabeled data.
To develop a deep-learning A.I. to assist with IVF, researchers at Brigham and Women's Hospital and Massachusetts General Hospital used thousands of embryo image examples to create a system that could differentiate and identify embryos with the greatest potential for success.
In fact, this deep-learning system performed considerably better than a group of 15 experienced embryologists. Coming from five different fertility centers across the United States, these embryologists underperformed when competing with the software to determine which embryos had the greatest chance of implanting out of a group of high-quality embryos.
The images used to train the system were of embryos at 113 hours post-insemination, and out of 732 embryos, the system had a 90% accuracy rate of choosing the highest-quality embryos. In general, the A.I. system had an evaluation accuracy of about 75% while the embryologists had an accuracy of about 67%.
Selecting Embryos: Then vs. Now
Before these advances in artificial intelligence, doctors and specialists had to rely primarily on observation and their own experience when selecting embryos to implant, and these methods led to lots of variability in the selection process. Because the expansion of these A.I. systems are still in progress, in some ways embryologists still need to rely on their personal skills.
However, the continued development of automated A.I. systems is promising for the future of IVF and is already offering increased benefits. With further expansions in this area of artificial intelligence, results of IVF treatment will continue to become more consistent.
Cost, Access, and the Future of A.I. in IVF
The current tools that are used to make decisions and evaluations about embryos are limited, subjective, and often expensive. As access and implementation of deep-learning A.I. systems increases throughout the field of IVF and fertility medicine, costs will decrease, process will be faster, and errors will be reduced.
With A.I. becoming increasingly prevalent in IVF, procedures like embryo selection will continue to become more precise. Through this precision, fewer IVF cycles will be required to reach a successful pregnancy, and the risks of IVF (such as multiple births) will also be minimized. Overall, these advances in A.I. technology indicate that IVF success rates will continue to improve over the coming years.
As with all areas of medicine and industry, advancements in technology lead to significant improvements and enhancements. Within the field of IVF treatment, advancements in artificial intelligence have simplified and standardized processes that have historically been costly and complicated. No longer needing to rely only on human judgement, embryologists and other IVF medical professionals can now increase the accuracy of their work with the assistance of strategically designed A.I. systems.
Country singer Granger Smith, and actress Amber Smith are currently expecting a baby. This exciting news for the couple and their family is all the more meaningful when considering their family history...
Catching up on Granger and Amber Smith and their family!
Scientists have found, created, and developed a new method to screen pregnant women for harmful prenatal environmental contaminants like air pollution using a DNA biomarker. These harmful prenatal environmental contaminants are linked to childhood illness and some developmental disorders...
A new test is being used to determine and predict problems with a fetus before it is born.
The Museum of Contemporary Photography (MoCP) has a new exhibit, and it centers around reproductive health. The exhibit is called “Reproductive: Health, Fertility, Agency.” The exhibit centers around artwork that is related to or comments on women’s rights...
Looking at the MoCP's new exhibit on reproductive justice and fertility.