AI, hormones and the future of fertility

Hormones play a crucial role in reproductive health, and advancements in artificial intelligence (AI) are transforming how we approach fertility treatments. In this blog Dr Ali Abbara, consultant in endocrinology, gives an insight into the research shaping the future of fertility treatments and women’s health.

What inspired your career in endocrinology?

I have always been fascinated by endocrinology as a specialty. Hormones have such a huge impact on most systems in the body and hormonal health is essential for wellbeing. Endocrinology is both beautifully logical and simple but also incredibly complex and nuanced at times, providing an exciting and challenging field of work.

Research is a particular passion of yours. Tell us about a recent study you were involved in which looked at the potential use of AI to personalise IVF treatment.

IVF treatment generates a large amount of complex data and AI is ideally suited to handling this complexity and providing insights that can enable us to optimise care. Many couples have limited access to funding for IVF treatment, and so getting the best possible outcome for each cycle is incredibly important.

During IVF treatment, the ovaries are stimulated with hormones so that follicles (the small sacs containing the eggs) grow to the right size for the next stage of treatment called the ‘trigger’ (which prepares the eggs for collection and fertilisation by sperm). If the follicles are too small or too large, they respond less well to the trigger.

So, deciding when to proceed is an important decision during IVF treatment and clinicians will closely monitor the growth of the follicles every one to two days using ultrasound scans. As there are lots of follicles of various sizes, often a rule-of-thumb is used based on only the largest follicles to simplify decision making, but this doesn’t consider the size of all follicles.

In our study, AI identified the chance of each follicle of a certain size being able to yield an egg. We showed that having more of the optimally sized follicles led to more eggs being retrieved and better live birth rates. This provides useful information for clinicians managing cycles to optimise clinical outcomes after each cycle and in future, we aim to develop a decision support system that can make the relevant recommendations.

Dr Ali Abarra
"Hormones have such a huge impact on most systems in the body and hormonal health is essential for wellbeing."

Where else do you think AI could have an impact within reproductive endocrinology?

There are several other avenues where AI could help us make more accurate decisions regarding diagnosis and treatment.

We are currently working on improving the diagnosis of menstrual disturbance, which can often signal when someone isn’t ovulating, and to help guide the selection of the best medication and dose for the restoration of ovulation. AI should enable us to make the most of the available data and optimise decision making.

Kisspeptin is a protein that is essential for reproductive hormone secretion and fertility. Based on your other research, how do you see kisspeptin improving IVF outcomes?

Our data suggests that kisspeptin reduces the chance of over-stimulating the ovaries which can lead to a dangerous complication called ovarian hyperstimulation syndrome (OHSS). Many of the treatment decisions in current IVF aim to minimise this risk, such as by using lower doses of stimulation, or by cancelling the cycle and not transferring the embryos straight away (called ‘Freeze all’).

However, all these approaches risk worsening clinical outcomes, such as not getting enough eggs, or taking longer to achieve pregnancy. Being able avoid this risk could therefore hugely simplify IVF protocols and optimise clinical outcomes. The next step is to conduct a clinical trial to demonstrate these benefits over currently used agents.

Dr Ali Abarra
"Women’s health has historically been neglected, especially with respect to research, and it is critically important that more effort is placed to improve care and research for women’s health needs."

You have also been exploring the role of the anti-Müllerian hormone (AMH) in understanding conditions like polycystic ovary syndrome (PCOS). Tell us more about this.

AMH is produced by the follicles in the ovaries. In general terms, women are born with all the follicles that they will have and gradually use them up until the ovaries run out, which often signifies the menopause. This concept is called ‘ovarian reserve’ and AMH is a useful blood test that reflects the ovarian reserve.

Women with PCOS have lots of follicles and often have a high ‘ovarian reserve’ and consequently a high AMH. Therefore, ‘high AMH’ has now been incorporated into the latest guidelines as a feature that can be used to diagnose PCOS.

Our previous work has shown that the higher the AMH, the more features of PCOS are present, so in the future it is possible that AMH could be used on its own to diagnose PCOS, but at present it is mainly used as just a single feature of PCOS, and other features like menstrual disturbance are also needed to reach a diagnosis of PCOS.

Why is this research important?

Women’s health has historically been neglected, especially with respect to research, and it is critically important that more effort is placed to improve care and research for women’s health needs.

There has been some progress to date – the government recently released a call-to-action in the form of the ‘Women’s Health Strategy’ to formulate a strategy to redress this inequality. And many funders now naturally expect research to be conducted in both sexes.

However, it is important that we all make more of an effort to improve care and conduct the research that women need to improve their wellbeing. It is also important that men advocate for more funding in women’s health research, mentor female colleagues, and challenge biases in medical research.