If you go to the hospital for a heart scan in the near future, it might be analysed by an artificial intelligence (AI) tool called the Singapore Heart Lesion Analyser or Sense. This new system will be tested at three public hospitals in Singapore—the National Heart Centre Singapore (NHCS), National University Hospital, and Tan Tock Seng Hospital—starting in the third quarter of 2025.
The year-long trial aims to see how well Sense works in real hospital settings. Around 300 patients will take part across the three sites. The system has been designed to reduce the time it takes to check for coronary artery disease. While doctors and radiographers usually need two to four hours to study a cardiac scan, Sense can do it in under 10 minutes.
That doesn’t mean doctors are out of the picture. Assistant Professor Lohendran Baskaran, a senior consultant at NHCS, explained that the system is there to support, not replace, clinical decisions. “Ultimately, all of this has to be reviewed, checked, and confirmed by the doctor,” he said. “This will never override a doctor’s position or clinical judgment.”
How Sense works and why it matters
Sense uses powerful computing and smart algorithms to study heart CT scans. It looks at calcium levels in the coronary arteries and fat around the heart—known as epicardial adipose tissue. These are key signs that help doctors determine whether someone has coronary artery disease, a condition in which the heart’s arteries become narrowed or blocked.
So far, the system has shown between 85% and 99% accuracy in controlled testing. However, this trial is important because it will show how Sense performs in real-world situations, where cases can be more complex and doctors manage many other tasks.
Coronary artery disease is serious—it caused nearly one in three deaths from cardiovascular problems in Singapore in 2023. That’s 8,311 people. Faster, more accurate diagnosis could help lower that number by catching issues early and guiding the proper treatment.
The lab and team behind the AI system
Sense was developed by the CardioVascular Systems Imaging and Artificial Intelligence (CVS.AI) lab at NHCS, working with the Institute for Infocomm Research under A*Star. CVS.AI was set up in 2021 and focuses on using AI to detect and predict heart problems faster and more accurately.
The lab covers a 164-square-metre area in the NHCS building on Hospital Boulevard. It’s fitted with advanced computing tools, including high-speed graphic processors that help handle huge amounts of data quickly. This setup lets the team build better AI models by training them on large, complex datasets.
Sense is built on earlier work by the same lab. That earlier system, Apollo, collected almost 3 million images of about 5,000 heart patients from CT scans over four years. It was designed to improve the speed and accuracy of CT scan interpretation. That data helped train Sense to become more accurate in diagnosing heart issues.
Professor Lohendran, who also leads CVS.AI, said the lab’s research will help Singapore better understand coronary artery disease. “We still don’t fully know which groups of people are most at risk or how severe the disease is in different parts of the population,” he explained.
This trial could make heart scan analysis faster and more efficient, potentially saving lives and reducing pressure on busy doctors.
