Cracking the wall' between AI and medicine

Artificial intelligence could revolutionise healthcare, but that doesn't mean you should rush headlong into it, says physician-scientist-innovator Wong Tien Yin.

- Tan Dawn Wei

Nowhere has the excitement over how artificial intelligence (AI) will transform our lives been more potent than in the area of medicine, where it's been held to such lofty standards that people expect it will some day—hopefully soon—wipe out epidemics, extend human lifespan and maybe even cure cancer.

The idea that AI will improve healthcare is not up for dispute. Real-life applications have already yielded positive results in more accurate and faster diagnostics, and streamlined operations in medical settings.

But the hype over AI as a cure-all that will replace doctors, save healthcare systems billions of dollars quickly or democratise access to high-quality care around the world needs a reality check.

In other words, don't be sucked in by the media amplification or AI companies overpromising their tech capabilities, or assume that narrow AI successes can widen to other areas.

Plenty of steps need to be taken before AI can work well for healthcare, and this doesn't just involve datasets and algorithms, says Professor Wong Tien Yin, who was medical director of the Singapore National Eye Centre from 2014 to 2021.

A pioneer in using digital technology in medicine, he was recognised in May for his work in this field and elected as a Fellow to the prestigious Royal Society, the oldest continuously existing scientific academy in the world that counts Charles Darwin, Albert Einstein and Stephen Hawking among its members.

The key is how to "crack the wall" between tech and medicine; something Prof Wong, one of the most-cited ophthalmologists in the world, had been doing long before AI became a mainstream fixation.

In November 2024, he and his colleagues published a case study in the New England Journal of Medicine on Singapore's experience in instituting the use of AI in eye screenings for diabetes, offering a how-to guide on integrating such innovations into clinical practice.