AI and microfluidics show promise in ART, but they must demonstrate benefit for patients and not just for clinics
AI, microfluidics and other emerging technologies offer a real opportunity to improve IVF practice, but clinics must ensure their use leads to meaningful outcomes for patients.
This was the message from an invited presentation on the future of reproductive medicine from Marcelle Cedars from the University of California, with insights into how lab-on-a-chip systems or machine learning could transform fertility labs through automation.
However, on the one hand, Professor Cedars, a past president of the ASRM, described AI and microfluidics as ‘the future’, but the other noted that these technologies do have downsides and should only be used with a clear objective in mind. She said: ‘These tools are going to be phenomenal. But we need to think about a meaningful outcome and the patient benefits, not to be enamoured with the next big new shiny thing. The procedure should be in the best interest of the patient, not about “did we get more MIIs”.’
Despite the huge advances in IVF, culminating today in longer acting medications, remote monitoring and the applications of AI, IVF still requires a lot of manual manipulation, she said, which means variability between labs and problems that need solving. The key question to ask though, she proposed, is what is the problem that can be solved through innovation.
She attributed the attraction of AI to its ability to process large amounts of data, with benefits which include predictions based on patterns invisible to the naked eye – for example, algorithms to predict ploidy status without the need for biopsy. Ovarian stimulation dosing and trigger timing, sperm selection and embryo selection are already well in development, but warned that ‘we don’t want to process an embryo with our own biases, which is why AI will thrive and change what we do’.
Focusing on the IVF lab, Cedars described microfluidics as another likely useful tool. There has been interest since the 1950s in how miniaturised systems could be deployed, and the technology has now emerged in fertility procedures such as sperm sorting and injection. She outlined the two types of microfluidics: mechanical which can help with ART procedures such as cell manipulation in ICSI, and biochemical such as the regulation of air quality.
The benefits of microfluidics, she explained, include rapid fluid manipulation, real-time monitoring, and a reduction in reagent consumption. With any new approach, there are downsides and microfluidics are no exception, said Cedars. She described the technology as complex, and fabrication and translation to mass production is a challenge. Prototypes are often made from materials that may be unsuitable for medical devices, hence mass manufacture is currently problematic.
Concerns also exist around the impact of microfluidics on the IVF process - media changes lor loss of embryo secreted factors. In addition, microchannels may be big enough for eggs and sperm but not for blastocysts – so prototypes would need to be modified. Microfluidics simulate complex physiological systems and Cedars stressed the important role of basic scientists in creating devices based on the technology.(1)
1. Ferraz MAMM, Rho HS, Hemerich D, et al. An oviduct-on-a-chip provides an enhanced in vitro environment for zygote genome reprogramming. Nature Comm 2018; 9: 4934.