GenAI: a new fertility planning tool?

Generative artificial intelligence (GenAI) – AI which creates new content by learning patterns and structures from vast amounts of existing data – is transforming healthcare in fields from oncology, radiology and cardiology to fertility planning. 

AI has been used in fertility since the late 1990s, first to predict IVF outcomes, then for sperm cell classification, oocyte selection, and embryo assessment, using machine learning and deep learning algorithms that analyse images and data. This AI-powered analysis of fertility ultrasound scans can help fertility specialists determine the optimum time for egg retrieval more accurately and far more rapidly than manual measurements. 

AI can also help embryologists make more accurate predictions about embryo viability and potential pregnancy success, potentially improving patient outcomes by identifying the best embryos for transfer, and informing treatment decisions. Today it is being used in leading South African fertility clinics, including Medfem, Aevitas and Vitalab. 

What are the benefits of GenAI for fertility treatment?

“The main advantage of big data analytics is the ability to accurately predict success rates for individual patients based on specific circumstances,” says Medfem psychologist Dr Mandy Rodrigues, a contributor to the European Fertility Society’s (EFS) Standards Development Group, which has produced a Fertility Patients Care Guide. 

The guide explains that “specialised technology platforms use databases containing thousands of patient profiles with very detailed information on medical history and fertility examinations, as well as outcome data. This data is used to find similarities and draw conclusions regarding potential patient outcomes, and to estimate individual success rates. 

“Building on big data analytics, AI can enhance technology platforms that estimate success rates of ART (assisted reproductive technologies) such as IVF (in vitro fertilisation). Machine learning and the accurate identification and interpretation of data patterns by AI systems can reach conclusions potentially beyond human capacity.”

AI contributes to better outcomes in ART by not only estimating success rates, but by also recommending treatment methods and protocols for the best chances of success, says Dr Rodrigues. “It can impact three key areas of fertility treatment: estimating outcome percentages more accurately; enabling better patient service, selecting appropriate patient protocols, and evaluating suitable types of treatment and lab methods to optimise results; and assessing cost-effectiveness of fertility treatments and reducing unnecessary expenses from failed cycles.”

What are the limitations of GenAI for fertility planning?

For all this, Dr Rodrigues says, AI has limitations: “It cannot replace the expertise and experience of qualified and caring fertility specialists and embryologists.”

Greg Tinney Crook, laboratory director at Aevitas, concurs. “We are acutely aware that the spark of life is not something we control,” he says. “But by using the best possible technology and science available, along with the extensive training and experience of our laboratory team, we are able to offer patients the most optimal embryos possible.”

Fertility treatment is highly individualised and complex, requiring clinical expertise and patient involvement, says prominent British embryologist Emma Whitney.  “For example, while AI can assess an egg or embryo’s visual characteristics, it can’t evaluate the many biological and genetic factors influencing its viability.” In addition, AI insights depend on the quality of the data inputted, and if an AI model draws on substandard or biased data, its recommendations can be misleading.

Can ChatGPT help you get pregnant?

Similar reservations are being raised amid the excitement around the growing use of ChatGPT in fertility planning. ChatGPT – AI focused on conversational interactions – is increasingly being used for clinical decision support, patient education, data analysis, and medical education. But just how reliable is it as a source of fertility information? 

A 2024 study in the journal Fertility and Sterility evaluated ChatGPT answers to questions about infertility and compared them to reputable resources, including the WHO (World Health Organization), the NIH (National Institutes of Health) and CDC (Centers for Disease Control). Overall, ChatGPT answers were deemed to be detailed and accurate.

“ChatGPT may serve as an easily accessible reliable source for fertility-related information,” the researchers concluded. However, they cautioned that it is “important to verify and fact-check information on infertility provided by ChatGPT”.  

“Always speak to your trusted healthcare provider,” says Pretoria obstetrician and gynaecologist Dr Elizabeth de Gouveia.

Links:

1. https://www.evewell.com/support/ai-fertility/

IMAGE: 123rf.com