A new artificial intelligence system, named "Sense" (Singapore Heart Lesion Analyzer), is poised to revolutionize cardiac care in Singapore. This system is designed to dramatically reduce the time required for analyzing cardiac scans, thereby enabling faster diagnosis of coronary artery disease. A yearlong trial of "Sense" is scheduled to commence in the third quarter of 2025, involving 300 patients across three prominent healthcare institutions: the National Heart Centre Singapore (NHCS), the National University Hospital, and Tan Tock Seng Hospital.

"Sense" leverages sophisticated computational capabilities and advanced algorithms to interpret cardiac imaging scans. Its primary function is to evaluate the risk of coronary artery disease in under 10 minutes. This represents a significant improvement over the conventional process, which, according to Assistant Professor Lohendran Baskaran, a senior consultant with the NHCS cardiology department, typically requires two to four hours of meticulous analysis by radiographers and cardiologists. Assistant Professor Baskaran also noted that this manual analysis can sometimes extend further due to the concurrent clinical responsibilities of medical professionals, such as attending to patients in clinics.

The AI system automates the critical process of analyzing CT scans. Specifically, "Sense" focuses on quantifying the amount of calcium deposits within the coronary arteries and assessing the epicardial adipose tissue, which is the layer of fat surrounding the heart and major coronary arteries. Initial evaluations conducted in controlled laboratory environments have indicated an accuracy rate for "Sense" ranging between 85 percent and 99 percent. However, Assistant Professor Baskaran emphasized that the upcoming yearlong trial across the three institutions will provide a more robust assessment of the system's accuracy and performance in real-world clinical settings.

It is crucial to note that "Sense" is intended to augment, not replace, the expertise of medical doctors. Assistant Professor Baskaran firmly stated, "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." This underscores the system's role as a powerful decision-support tool.

The development of "Sense" is spearheaded by the CardioVascular Systems Imaging and Artificial Intelligence (CVS.AI) research laboratory, based at the NHCS. This innovative system builds upon the foundations laid by an earlier CVS.AI initiative known as "Apollo." The "Apollo" project was an AI-driven national platform for CT coronary angiography for clinical and industrial applications. Over a period of four years, "Apollo" successfully compiled an extensive database comprising nearly 3 million images derived from the CT scans of approximately 5,000 cardiac patients in Singapore. This repository also includes comprehensive clinical data. Zhong Liang, co-director and core technical lead of CVS.AI, highlighted that this substantial dataset is instrumental in enhancing and strengthening the AI algorithms through the application of big data analytics.

The development of tools like "Sense" is critical given the global burden of coronary heart disease. It is estimated that over 250 million people worldwide are living with this condition, affecting approximately 145 million men and 110 million women. Coronary heart disease is a major cause of mortality, responsible for an estimated 9 million deaths annually. According to the British Heart Foundation, it was the world's single biggest killer in 2021. In Singapore, coronary artery disease contributes to nearly one-third of all cardiovascular-related deaths. In the year 2023, 8,311 individuals in Singapore died from cardiovascular disease, which accounted for about 30 percent of all deaths in that year. Zhong Liang also pointed out that certain aspects, such as the severity of the disease and which segments of the population are most likely to be affected, are still not fully understood, indicating areas for ongoing research.