Faster AI detector of cardiac disease to be trialed in Singapore

A new artificial intelligence system which will reduce the time required for a cardiac scan analysis and allow for faster diagnosis of coronary artery disease will be available in Singapore as part of a yearlong trial.

The Singapore Heart Lesion Analyzer, or "Sense", will be trialed at the National Heart Centre Singapore, or NHCS, the National University Hospital and Tan Tock Seng Hospital in the third quarter of 2025.

A total of 300 patients across the three institutions will be involved in the trial.

"Sense" will use sophisticated computational capabilities and algorithms to interpret cardiac imaging scans and evaluate the risk of coronary artery disease in under 10 minutes, a dramatic improvement from the original several hours.

Originally, this is a process which typically requires two to four hours of analysis by radiographers and cardiologists, said Assistant Professor Lohendran Baskaran, senior consultant with the NHCS cardiology department.

"Sometimes it can take longer because doctors are busy doing other things in between all these scans, such as seeing patients in-clinic," said Lohendran.

"Sense" automates this process, analyzing CT scans to establish the amount of calcium deposits in the coronary arteries and epicardial adipose tissue — the layer of fat surrounding the heart and major coronary arteries.

It has shown between 85 percent and 99 percent accuracy so far, Lohendran said, though he noted these results have been from the use of "Sense" in a controlled environment.

The yearlong trial at the three institutions will give a better idea of how accurate the system is in real-world settings, he said.

"Ultimately, all of this has to be reviewed, checked and confirmed by the doctor before taking it any further. This will never override a doctor's position or clinical judgment," said Lohendran.

Big data technology

"Sense" is being led by the CardioVascular Systems Imaging and Artificial Intelligence, or CVS.AI, research laboratory at the NHCS. It builds on an earlier project by CVS.AI — the AI-driven national platform for CT coronary angiography for clinical and industrial applications, or Apollo.

Over four years, it built a database of almost 3 million images from the CT scans of some 5,000 cardiac patients here, together with comprehensive clinical data.

The database will allow CVS.AI to enhance and strengthen its AI algorithms, using the big data collected and analyzed, said Zhong Liang, CVS.AI's co-director and core technical lead.

Top killer

It's estimated over 250 million people are living with coronary heart disease. Globally around 145 million men and 110 million women have coronary heart disease.

Coronary heart disease kills an estimated 9 million people each year — in 2021 it was the world's single biggest killer, according to British Heart Foundation.

Coronary artery disease is responsible for almost a third of the cardiovascular-related deaths in Singapore.

In 2023, 8,311 people died from cardiovascular disease in Singapore, accounting for about 30 percent of all deaths that year.

Factors such as the severity of disease and what part of the population is likely to have it are still poorly understood, Zhong said.

The Straits Times, Singapore