SGH working with A*STAR to commercialise AI-powered antibiotics test for superbugs

SINGAPORE - Singapore General Hospital (SGH) is working with A*STAR’s Diagnostics Development (DxD) Hub to commercialise a test kit that helps doctors determine which combinations of antibiotics work best against drug-resistant infections.

Once its production is scaled, the test kit, which has been used 84 times in the past three years at SGH, can be rolled out to other hospitals and laboratories locally and overseas.

The two organisations’ collaboration is part of a three-year memorandum of understanding signed at SingHealth’s AI in Health x ATx26 Symposium held on May 19.

The goal of the collaboration is to create diagnostics technologies that can be licensed or spun off to local or foreign companies, said Dr Bernard Ng, director of innovation, enterprise, and commercialisation at SGH.

“One of the ways to ensure that our projects impact more patients is through commercialisation. This is also aligned with our national agenda of caring for patients while creating economic value,” said Dr Ng. Licensing and royalty fees from the commercialisation of the technology will go towards funding future innovation projects at the hospital, he added.

The test kit for combinations of antibiotics is broth-based and must be kept frozen at -80 deg C, which makes export and logistics difficult.

So SGH is working with A*STAR’s DxD Hub to create dry versions of the test kit to allow its export to potential markets like South-east Asia, Hong Kong, and the Greater Bay area in China.

Called the Individualised Antibiotic Combination Test, or iACT, the biotechnological innovation contains 180 different proprietary combinations of existing antibiotics developed over more than 15 years by SGH infectious diseases pharmacists and clinicians.

Each combination mixes different drugs in varying pairings and concentrations to determine which works best against a patient’s drug-resistant bacteria.

After the bacteria is tested against the 180 combinations, AI-powered predictive software analyses the results alongside the patient’s medical history to generate a personalised report recommending the most effective antibiotic combinations.

Antimicrobial resistance is a growing global health threat that makes common infections harder to treat, leading to longer hospital stays, higher medical costs and more deaths.

Immunocompromised patients are more vulnerable to anti-microbial infections because they tend to develop repeated infections and require frequent antibiotic treatment, giving bacteria more opportunities to evolve resistance.

“Without this test, doctors may have to use last‑resort antibiotics that can seriously damage the kidneys. With our combinations, we can often avoid those drugs and still control the resistant infection,” said Associate Professor Andrea Kwa, deputy director of pharmacy (research and innovation) at SGH.

Other projects that could be commercialised include PENSIEVE-AI, an artificial intelligence-powered drawing test developed by SGH and Singapore’s Government Technology Agency (GovTech) that can detect early signs of pre-dementia in under five minutes by analysing how seniors complete a series of drawing tasks on a tablet.

Another agreement signed at the symposium was between public healthcare cluster SingHealth and the Gyalpozhing College of Information Technology, an institution under the Royal University of Bhutan.

Under a two-year memorandum of understanding, the two institutions will work together to develop an AI-assisted chest radiograph model to help Bhutanese clinicians, especially in rural hospitals with limited specialist expertise, detect lung diseases such as infections and cancer more quickly and accurately from chest X-rays.

The project builds on Singapore-developed AI foundation model MerMED-FM, created by SingHealth Duke-NUS Academic Medical Centre and A*STAR Institute of High Performance Computing, which was trained to analyse a wide range of medical images and diseases.

The model will be trained using data from Bhutanese patients, and is expected to roll out in 2027 at three hospitals in Bhutan.

SingHealth and the Gyalpozhing College of Information Technology will also jointly develop guidelines, educational programmes, and regulatory frameworks on responsible AI use in healthcare, tailored to Bhutan’s context.

When it comes to AI in healthcare, Singapore is starting from a very good position, with a strong healthcare system, the right talent, and sound governance, said Minister of State for Health Rahayu Mahzam.

“Healthcare is one of our national AI missions, and there is so much potential for AI to create greater impact in healthcare – to improve clinical decision-making and healthcare delivery and helping people to live longer and live well,” said Madam Rahayu, who is also Minister of State for Digital Development and Information.