Skip to content

Envision Medical Imaging 29 September 2021

Early detection of heart disease

Dr Brendan Adler

Early detection of heart disease

By Dr Brendan Adler, CEO Envision Medical Imaging

"The good physician treats the disease; the great physician treats the patient who has the disease." – Sir William Osler.

The leading cause of death in Australia is heart disease, and for a significant proportion of those people, their first presentation of heart disease is sudden and often deadly. However, we have many early detection programs for other diseases such as breast cancer and colon cancer that have been proven to save lives.

Sudden death from heart disease more commonly affects younger working-age people in their 50s and 60s than cancers and other chronic diseases. Nevertheless, about half the population will die of coronary vascular disease.

At present, our best way to measure risk is to use "risk factors" such as blood pressure, diabetes and cholesterol levels to predict the chance of subsequent heart attacks in the population. Studies show that risk factors are present in most who have a heart attack. The problem is that they are also present in most who do not have a heart attack. Therefore, they are not very powerful at predicting individual risk.

Fundamentally, risk factors are a guess. A best guess, but a guess about individual risk. They are used because, historically, there has not been a way to visualize the actual disease process. In the same we use a mammogram to "look" for early breast cancer, the development of coronary artery CT scans now allows us to easily visualize the disease (atherosclerosis) and its potential complications (narrowings) of the coronary arteries.

Trials show that coronary event prediction by CT is much more robust on its own than risk factor scores, which is additive to risk factor scores. This makes intuitive sense as risk factors define risk, which is not the same as "seeing" the disease". Who would you think was more at risk? An elderly man with high cholesterol or a young woman with a significantly abnormal CT scan? It isn't the elderly man.

Risk factors are very useful to treat if the patient has the disease; lowering cholesterol, for example, is of known benefit in those with coronary artery disease. So in principle, the bigger picture needs to be, firstly define the disease, and secondly treat the risks.

Regardless of benefit, the current cost to look at everyone's coronary arteries with CT would be prohibitively expensive. When would you look? How often? If a case were made to the government to review the cost-benefit analysis of screening for life-threatening coronary artery disease, it would likely be rejected on this basis.

However, if technology could decrease the cost and difficulty of reviewing and reporting coronary CT angiograms, widespread use to detect treatable disease in high-risk individuals would be more affordable, and potentially the cost-benefit for the wider population could be significant.

There has been extensive research into, and now development of, efficient technologies. Artrya is one company that is leading the way in AI detection of coronary artery disease here in Australia, collaborating with Envision Medical Imaging, The University of Western Australia, Harry Perkins Institute of Medical Research, Professor Girish Dwivedi of Fiona Stanley hospital, and The University of Ottawa Heart Institute, Canada.

These technologies obtain fast and effective analysis of coronary CT scans using AI algorithms in minutes to create images and reports that would ordinarily require significant time by current imaging technologists. This AI approach may allow clinicians to see many more patients at a significantly lower cost. There is a future potential opportunity to make a societal impact with earlier detection, should governments, payers and health systems be willing to change patient care models.

My hope is that we start the conversation around making early detection and treatment of heart disease an affordable and accessible option for future adoption. Our goal should be to focus on treating patients with early coronary artery disease, not symptomatic late disease, and with new AI technologies now in the clinical domain, there is a real step forward in the cost-benefit of disease detection on our doorstep.