The number of Swedes suffering from acute myocardial infarction is decreasing, but it still stands at over 20,000 per year. Johan Malmqvist has led a research team that has investigated how to identify patients at risk as precisely as possible.
In the study, the research team analysed coronary X-ray images from 23,000 healthy people in early middle age and followed their health. The participants were included in the large population study Scapis.
Finds with more
The researchers were able to determine that a new Swedish AI software that reads X-ray images is better than the human eye at detecting harmful plaque in the coronary arteries.
In Scapis, about 40 percent of people have some plaque that can be seen with the naked eye. Our AI algorithm found plaque in about 80 percent of the participants, says Johan Malmqvist.
The researchers were also able to determine that the plaque the AI model detected, but which the human eye could not see, meant an increased risk of infarction. This was shown by follow-up health data on those who later developed the disease.
It has not previously been known that small AI-detected plaques are significant. So the study provides further evidence that AI adds valuable information, says Johan Malmqvist.
Appropriate medicine
The study shows, above all, that with AI support, it is possible to obtain more information about patients who have already undergone coronary angiography. AI-supported image interpretation can also provide doctors and patients with better information about appropriate medication.
The study does not answer the question of whether to perform a chest X-ray on patients who do not have symptoms; other studies are needed for that.
But this is an important piece of the puzzle in how to best assess the risk of future heart attacks, says Johan Malmqvist.
The study is published in the journal JACC Cardiovascular Imaging.
In 2024, just over 20,000 cases of acute myocardial infarction were reported. Approximately 4,000 people died from heart attacks.
The research initiative Scapis (Swedish Cardiopulmonary bioimage study) aims to identify people at risk of cardiopulmonary disease in a timely manner. Around 30,000 participants between the ages of 50 and 64 have taken part and undergone extensive examinations.
A number of studies have been conducted using data from Scapis. One of these is the one on AI-based interpretation of X-ray images.
The Swedish Heart and Lung Foundation is the main financier of Scapis.
Sources: National Board of Health and Welfare and the Swedish Heart and Lung Foundation
