From predictive medicine to image analysis automation, AI is playing a vital role in our care. But some fields of healthcare have been slower to innovate, and patients are suffering directly as a result.
AI is no different from any other new, disruptive innovation. There’s a tried-and-true formula for success and adoption. As with any new technology, an adoption cycle must take place. Early adopters dive in and tinker with new gadgets, whereas mass adoption takes time.
Think of the automobile. People needed to understand what the car was all about. Then they had to trust that the car was worth the price – and would be a safe mode of transportation.
Education and trust. It was the case with cars. And it is the case with AI.
Fertility Care Needs AI
90 million couples globally, and one in five (~19 percent) women in the United States alone, experience fertility challenges. Even with treatment advancements, only 30 percent of prospective parents have successful IVF outcomes in one treatment cycle. Percentages drop as women age. Shockingly, 48% of couples with difficulty conceiving don’t consider their condition to be infertility.
Given how major the issue is, the rapid embrace of digitization and AI is essential to improving these figures and supporting the IVF industry in better serving their patients.
These six areas can change everything for IVF. We should consider each of them as we move from utilizing limited assessment parameters that can only measure, to an extent, the probability of IVF success, towards a data-driven treatment process that more accurately informs the end-to-end IVF journey.
-
Access to IVF Should Improve…and It Will
This is both an imperative and an expected result. Globally, IVF treatment access is too limited for prospective parents. While treatment demand over the past 10+ years has doubled, the number of clinics has grown by less than five percent.
Today, too much time is spent by too few IVF professionals on routine tasks that can either be automated or conducted by non-specialists. Without full adoption of AI tools to automate routine processes in IVF clinics, embryologists and fertility specialists are simply unable to meet the demand.
If we put data and AI at the center of IVF, these IVF professionals can handle far more cycles per week and see pregnancies taking place with fewer IVF cycles per patient.
The result will be an increase in availability of time and resources for embryologists and REIs, and a significant reduction in the cost-per-case. This means a better financial model for all three key players in the process: prospective parents, IVF professionals and payers.
This can only be the case if…
-
IVF Professionals Understand the True Promise of AI
The large volumes and visual complexity of data captured throughout the IVF process make it a great candidate for effective AI analysis. But current machine learning solutions have not illustrated enough of a potential impact, evidenced by its limited adoption in IVF.
As innovators, we must take on the burden of market education to ensure IVF professionals understand what AI is, particular to their work, explaining where it can (and can’t) help and how to interpret results to best support decision making.
However, given our (innovators) vested interest in the success of our innovations, our voice alone is not enough.
There are certain IVF professionals – both medical and laboratory staff – who are championing such progress. Their viewpoints should be amplified for the sake of education.
-
True Partnership Between IVF Clinician and Tech Requires Transparency
Once IVF professionals understand the opportunity and limitations of AI, they can begin assessing how AI can best work for them.
For instance, an AI system providing a quality score on an ovum, sperm (for use in ICSI-based IVF) or embryo with no further explanation of how it arrived at the score will almost always lead the IVF professional to err on the side of their own experience.
True partnership between IVF professionals and technology will only be possible when AI results are explainable. This level of transparency is essential for IVF teams to feel comfortable adopting AI and will also naturally foster greater transparency with their patients.
-
Patients Deserve Their Data
For decades, the IVF world was unable to access the data generated from genetic testing and analysis of embryos. Rather, IVF professionals and patients alike had to depends solely on IVF professionals’ experience and knowledge concerning embryo viability.
Today, patients have the right to own their health data and want to take a more active role in treatment decisions. They are requesting data as detailed as genetic screening analysis for each embryo, so they can understand their particular journey better.
Simply put, patients are paying more attention and want to be more involved.
Importantly, IVF professionals want to share this information. But they need access to the data to provide the transparency that patients seek. Explainable AI-derived data is the answer that IVF clinicians have needed to open this conversation and provide clarity to patients at the most granular level.
-
Credible Clinical Evidence Must Be Provided for AI to be Accepted by the System
A cornerstone to adoption of healthcare solutions is clinical evidence – an area that has been oversimplified in IVF. Companies are investing in research, but embryo selection specifically, is an extremely complex process with many moving parts; any one part can be the difference between success or failure. As such, much of the AI-based tool research has focused on the ability of the software to meet the standards of current embryo selection processes.
For clinical research to earn credibility – and the trust of the medical community – it must be conducted prospectively, as randomized, blinded studies in real-world settings. This will demonstrate the viability of AI in treating patients and illustrate to the health ecosystem that such startups are holding themselves to a higher industry standard and are invested for the long haul.
-
The Pursuit of Regulatory Clearance
AI-based IVF platforms must show this respect to health ecosystem and its standards for clinical research. We can do this by embarking on the formidable task of seeking clearance by regulatory bodies responsible for protecting the safety of patients.
As innovators we know the potential power of our solutions. But we must be able to prove that our products are as effective as we claim. Whether it’s CE approval in the European Union or FDA clearance, securing these regulatory approvals is important to demonstrating compliance with established and expected standards of safety and care.
The Way Forward
We are on the cusp of a dramatic transformation in the IVF industry, as AI and data can help more prospective parents see their dreams come true.
To make this happen, we must commit ourselves to educating the health ecosystem about the legitimacy of utilizing AI solutions to improve patient outcomes. Moreover, we have to take ownership of improving transparency with patients and appreciate that increased access to IVF for all who seek it is possible.
The key to more affordable and more impactful IVF is maximizing what AI and advanced technology have to offer, while also upholding the standards that have served the health ecosystem so well for decades.
By Eran Eshed, CEO, Fairtility