NUH doctors use AI to identify stroke patients within a minute to give timely life-saving treatment

The RapidAI tool processes the imaging data from CT scans in less than a minute, whereas a manual scan needs up to 20 minutes. PHOTO: NUH

SINGAPORE – Doctors at the National University Hospital (NUH) are using an artificial intelligence (AI) triage tool to identify stroke patients in less than a minute and give them timely treatment, shaving minutes off the medical response to an emergency which could lead to severe disability, or even death.

The RapidAI tool expedites the process of identifying an acute ischaemic stroke, where a major blood vessel in the patient’s brain is blocked by a clot, depriving the brain of oxygen and nutrients. This leads to the death of brain tissue.

Endovascular treatment, which “unblocks” the vessel and restores oxygen flow to brain tissue, is a highly time-sensitive treatment. Administering it to the patient as soon as possible is crucial for the best possible outcome.

Dr Leonard Yeo, senior consultant in the Division of Neurology at NUH’s Department of Medicine, said: “With each minute of delay in opening a blocked artery, the patient could lose 1.9 million brain cells, hence ‘time is brain’. The earlier we unblock the artery, the better the patient outcome.”

Acute ischaemic stroke is the biggest cause of disability in Singapore and across the world. Before the use of AI, the process of identifying such stroke victims could take up to 20 minutes.

NUH started using RapidAI in February 2023 on suspected stroke patients at the hospital’s emergency department. Patients do not need to pay extra for its use.

More than 400 patients have been assessed with the tool and an estimated 30 to 35 of them received endovascular treatment. 

When a suspected stroke patient first enters the emergency department at NUH, he undergoes an emergency computerised tomography (CT) scan. The imaging data from the scan automatically flows into and is processed by the RapidAI platform. The tool then generates easy-to-interpret colour-coded images and sends the results via a secured e-mail to a medical team of stroke neurologists and interventional radiologists.

The entire process takes less than a minute, and includes valuable information about the proportion of the brain that has been damaged in the stroke, and the parts of the brain that can potentially be saved through medical intervention.

The information helps the team make a swift decision to carry out the most appropriate treatment.

Said Dr Yeo: “NUH is using the power of AI to shave life-saving minutes off an emergency like a stroke. This can avoid the extent of brain damage that can cause paralysis and loss of speech and cognitive functions.”

In the past, radiographers manually processed the CT scans. Doctors then logged in to their laptops or went to an imaging workstation to review the information.

Adjunct Associate Professor Anil Gopinathan, head and senior consultant in the Division of Interventional Radiology at the Department of Diagnostic Imaging at NUH, said: “Before we adopted this AI platform, it was impossible to expect all this information to be made so promptly available for the senior clinicians to review and make a decision.

“AI has the potential to completely change the speed at which therapeutic decisions are made, thereby reducing the amount of brain cells lost and improving the likelihood of a stroke patient walking out of the hospital.”

Early intervention was a key factor in Mr Tham Mow Cheong’s recovery from the stroke that he suffered in December 2022. While at work, the 68-year-old mechanical technician experienced dizziness. As he has high blood pressure, he asked his colleagues to take him to Ng Teng Fong General Hospital. 

He was transferred to NUH, where he was diagnosed with acute ischaemic stroke. Thanks to the early identification of his condition and timely treatment, he recovered well, recovering his motor functions and returning to work after the rehabilitation.

Stroke patient Tham Mow Cheong (centre) with Dr Leonard Yeo (left) and Adjunct Associate Professor Anil Gopinathan. ST PHOTO: JASON QUAH

Although Mr Tham was not diagnosed using the AI tool, his case illustrates how early identification can lead to the best recovery outcome.

Stroke rehabilitation typically takes up to three to six months. Early intervention plays a part in shortening the process. A 10-minute delay in treatment can add two to three weeks to the rehabilitation. 

In Singapore, stroke is the fourth leading cause of death, with a prevalence of 4 per cent among adults aged 50 and above. The stroke team at NUH manages more than 1,000 ischaemic stroke patients annually, with one in six cases involving endovascular treatment.

Dr Yeo said that NUH has set itself a target of getting at least half the patients with acute ischaemic stroke to receive appropriate treatment within an hour of arriving. He believes NUH can significantly improve on this target and shorten the time taken for patients to receive treatment with the use of RapidAI.

Join ST's Telegram channel and get the latest breaking news delivered to you.