In Singapore's healthcare revolution, AI is the key
By Nurfilzah Rohaidi
There needs to be productive discussions around experimentation in healthcare as AI becomes more prevalent, says Steve Leonard, CEO of SGInnovate.
Image: National Healthcare Group
New technologies such as artificial intelligence could prove “hugely important to humanity”, says Steve Leonard, Chief Executive of SGInnovate, a government-owned company that funds deep tech startups. He believes that Singapore is poised on the cusp of a healthcare AI revolution; in the near future, intelligent digital tools will help doctors and nurses speed up diagnoses and pick out trends and patterns.
On the sidelines of the GREAT Festival of Innovation in Hong Kong in March, Leonard discusses the cutting-edge of AI in healthcare, but also stresses the need for people to trust and accept the technology.
An AI toolkit
As its population ages, Singapore is putting greater focus on healthcare innovation, and reconsidering how it approaches healthcare management. AI’s value for healthcare lies in its ability to parse through and analyse millions and millions of pieces of data, Leonard explains. “This is about how AI can be a set of tools that can assist or help doctors provide care.”
SGInnovate bridges the gap between research and reality in Singapore. Leonard works closely with researchers and startups to bring great ideas and concepts from the lab bench to market; in other words, acting as “the commercial guys”.
“It doesn’t mean that these things are perfect, but we can’t keep them wrapped up in a box and never actually use them,” Leonard remarks. “We have to be ready to let things be experimented with.”
“We have to be ready to let things be experimented with.”Within SGInnovate’s walls, the true potential of healthcare AI is taking hold: two startups are using computer vision and AI to help doctors improve the accuracy of disease diagnosis - and in some cases, the tech itself does the diagnosing, Leonard continues.
One of these startups is developing “the stethoscope of the future”, which will allow doctors to ‘see’ into a body. The tool, which employs computer vision and AI, will have been trained to know what to look out for by learning from millions of images - and provide analyses accordingly, Leonard explains. “It's not replacing the doctor - it's giving a tool to him; just like the stethoscope didn't replace the doctor.”
Another is working on a solution that can predict the risk of relapse in stroke patients. It uses computer vision and fluid dynamics to measure the speed of blood flow in arteries and veins, Leonard says. While statistics show that one in four stroke patients are at risk of relapse, if it is possible to pinpoint patients with even greater accuracy, then they could receive treatment even before it happens, he points out.
“Every year that goes by, it’ll be more and more accurate; a lot of these computer vision algorithms are already at the level of accuracy of a professional radiologist,” he adds.
Both solutions are currently in beta testing at a few hospitals in Singapore, and it is entirely possible that in a few years, they could feature in “the toolkit of a GP”, Leonard remarks.
The impact
Healthcare bills are rising for both patients and the government. Singapore plans to spend S$10.2 billion on healthcare this year - more than double the figure in 2010 (S$4 billion).
Meanwhile, in a recent report, insurers cited Singapore as fourth highest in the region in terms of medical inflation costs, the Straits Times reported. Inpatient costs were identified as the main driver of medical inflation, the report said.
Innovations in healthcare tech could mean that limited resources are used more wisely, Leonard notes. “As people get older, they need more healthcare. It’s only going to become more of a caseload [for healthcare providers],” he remarks. “It is about expanding and adding to their capability.”
For instance, a Singapore-based surgeon was able to reduce his patients’ recovery period by using robotics in his surgery, Leonard explains, freeing up beds for other patients that need them. “He is able to get patients out the same day, when it used to take four days for them to recover,” he says. “Think about it - four days in the hospital, versus zero.”
Beyond the clinics and operating theatres, AI could prove very valuable in healthcare research. Millions of studies, papers and reports are produced every year, much more than any human can read, Leonard says. AI could be a “huge advantage”, helping to “distill this and bring a condensed version of it to the doctor”, he explains.
AI could also create real change in more remote regions, where a doctor may be out of reach. One of SGInnovate’s partner startups is developing a handheld device that can capture ultrasound images of patients in rural areas. The algorithm will be able to compare this ultrasound to millions of others, and give a recommendation for treatment, Leonard explains.
Changing perceptions
It is easy to get swept up in the excitement around AI - but Leonard admits that “social adoption is a bigger challenge than the technology”. There is still a discomfort that exists around its use in clinical diagnoses, he says. “Sometimes people would say, oh, that sounds scary, I want my GP to make that decision, don’t tell me that's some robot or AI,” he explains.
“Social adoption is a bigger challenge than the technology.”This lack of trust in AI is visible from people’s negative perceptions of autonomous vehicles - made worse by the high-profile death of a woman by a driverless Tesla car in Arizona in March. Autonomous vehicles could transform urban mobility in countries as dense as Singapore, but innovation in this area would not “be worth much” if users are nervous of this technology, Leonard points out.
SGInnovate hosts frequent dialogues and events at its headquarters in Singapore’s Chinatown, where “we just try to talk about it” and shift some of these perceptions, he says. These dialogues with industry and the public help to tease out some of the underlying concerns around humans’ relationships with emerging technologies, and raise awareness of the need for experimentation so that they can mature.
As one solution, the government has recently launched regulatory sandboxes for companies to experiment with health tech. It would allow companies to take risks with new services in a “safe” environment, without fear of regulatory backlash.
As demographics rapidly change, AI could be the key to managing healthcare challenges as they come. In a few short years, doctors could count digital tools as indispensable as their physical ones.