Singapore to train AI models using local clinical data, medical guidelines

Caption language: en · Fetched: 2026-07-12

Patients in Singapore could soon get diagnosed and treated with the support of AI models trained on local clinical data and medical guidelines instead of those from overseas. Health Minister Ong Ye Kung says the models will first focus on conditions such as diabetes and eye diseases before being rolled out across the healthcare system. Caitlin Ong with this report. >> Now for some AI could look like chatbots and office productivity, but here it's taking on a little bit more of a physical form. Now this is a robotic exoskeleton and it's designed to help accelerate recovery like in the case of stroke patients. It can help patients walk up to 15,000 steps in a single rehabilitation session, about three times what they might typically manage with a physiotherapist.

And because this exoskeleton can be operated by trained staff instead of physiotherapists, it can help alleviate pressure on an already limited workforce. And while the exoskeleton itself isn't necessarily powered by AI, a partnership between NCS and Forerun Rehab is building the software layer behind it. Now that's where the AI comes in. With pulled data, therapy settings can be tailored to a patient's needs and better help to monitor their recovery. >> Speaking at the NCS event where such tech was showcased, Health Minister Ong Ye Kung says most AI use cases in healthcare are ground-up initiatives. He says one promising initiative is a Singapore Medical Foundation AI model called Symphony. >> Symphony will build AI models that could serve as trusted companions to our clinicians.

They can suggest possible diagnosis, treatment pathways, and next steps, supporting but not replacing clinical decision-making. Such models already exist, but they were trained using patient data and medical guidelines in other countries. Their are will be developed and trained using Singapore's clinical practice guidelines and local clinical data contextualized to our population and eventually integrated into our clinical workflows. >> Healthcare is also one of several sectors where NCS is looking to scale AI adoption. Education and transport are among the others where it says the challenge is no longer experimenting with AI but deploying it at scale to transform the way organizations operate. >> 95% of our pilots, right, as has been reported, have never really scaled into production.

But we are a lot more optimistic at this point in time because we are seeing a lot more techniques, right, and large language models that are actually able to be applied into solution, right, and with very good and clear business objective, right, we are able to scale these AI solution into production. >> To support that shift, NCS is expanding its enterprise AI toolkit, including new platforms that help organizations build AI applications securely within their own environments. It also released a playbook drawing lessons from more than 100 projects to help businesses avoid common pitfalls when deploying AI. To train the next generation of AI talent, NCS is partnering Digital Industry Singapore to hire more than 130 AI practitioners over the next 3 years alongside new programs with universities and industry.