AI medical diagnosis start-up sets its heart on Singapore launch

The Straits Times by SEOW BEI YI

Singaporeans suffering from a heart condition may soon be able to get specialist help more quickly if a new artificial intelligence-driven start-up gains regulatory approval to distribute its system here.

Called Tricog, the Singapore-based company, which has received US$4 million (S$5.5 million) in funding to date, said its algorithm can provide patients with a diagnosis in six minutes based on their electrocardiography (ECG) data.

An ECG machine with a Tricog add-on encrypts a patient's results and sends them to a cloud, where a machine-learning algorithm interprets the data and forms a diagnosis.

This is verified round the clock by a team of 20 doctors, who help to train the system's accuracy, and the closest hospital is alerted so doctors can be on standby to provide treatment, said Tricog's founder, Dr Charit Bhograj.

Dr Charit, who is an interventional cardiologist, added that the service is available in 12 countries including India, Indonesia and Malaysia, and has diagnosed 1.5 million patients over the past 11/2 years.

"The idea is to amplify the work done by a few doctors, to take care of a lot more patients," he told The Straits Times. "Of the eight million people who die every year of heart disease, 50 per cent... can be saved if the heart attack is diagnosed quickly and they are treated.


Of the eight million people who die every year of heart disease, 50 per cent... can be saved if the heart attack is diagnosed quickly and they are treated. Diagnosing someone with a heart disease is where the gap is.

DR CHARIT BHOGRAJ, Tricog's founder.

"Diagnosing someone with a heart disease is where the gap is," he said, adding that to do so involves making sense of an ECG reading, which requires a specialist.

In other parts of the world, it may take up to six hours before a person with chest pains enters a hospital, he said, although intervention should ideally be taken in an hour or so.

Dr Charit's company is in talks to partner a Singapore maker of ECG machines, allowing its AI system to be integrated with these machines and used here.

He envisions the machines being used in polyclinics or at general practitioner (GP) clinics, so patients who seek help near their homes can be referred to a specialist more swiftly.

Tricog is also pending regulatory approval to start operations in Singapore and aims to do so next year.

On average, the National Heart Centre Singapore sees one heart attack patient each day. There were about 60 deaths a year on average in the past few years as a result of heart attacks and strokes that were possibly worsened by strenuous activities at work.

Dr Lim Choon Pin, a cardiologist at Mount Elizabeth Novena Hospital, said: "With an ageing population and increase in heart disease burden, there will likely be an increasing demand for technology that aids in remote monitoring and diagnosis."

He added that such devices also help to keep patients within their communities, as most prefer staying out of hospitals.

But Dr Lim said cloud-connected ECG systems are already in use here, adding that he uses one for patients suspected of having heart rhythm problems.

In most cases, he said, the ECG is normal when a patient consults his family doctor or cardiologist. Cloud-connected monitoring helps detect abnormalities if they happen during a patient's daily activities.

Dr Philip Wong, a senior consultant at National Heart Centre Singapore, said that while response time - from when a patient arrives at an emergency room to when his artery is opened - has been improving, it is unclear how long a patient may take to arrive at a hospital from a GP.

Tricog's system would help get patients to a doctor more quickly for intervention, Dr Wong said.

The cardiologist added that many GPs may not have ECG machines as the readings require a specialist's interpretation.

The machines in polyclinics are also not connected to a cloud for remote diagnosis.

But Dr Wong - who is also the founder of a medical technology firm whose Spyder ECG device monitors a patient's heart rhythm continuously and stores it in a cloud database for the patient's doctors to access - cautioned that there will be challenges in implementing Tricog here, such as the security of data on a cloud database.

Dr Lim said legal aspects need to be addressed as well: "If an abnormal rhythm is missed and an unfortunate event occurs, who is responsible?"