Researchers at the University of Oxford announced on Apr. 15 that they have developed an artificial intelligence tool capable of predicting heart failure up to five years before symptoms appear. The tool analyzes routine chest CT scans, which are commonly ordered for chest pain, and identifies signs of risk that are not visible to the human eye.
The new technology could provide patients and doctors with crucial early warnings, allowing more time for preventive care and potentially reducing hospital admissions. According to the university’s findings, changes in heart fat can serve as a sensor for impending heart failure.
The study, published in the Journal of the American College of Cardiology, found that higher-risk patients had a one-in-four chance of developing heart failure within five years. The AI system predicted this outcome with 86% accuracy after analyzing inflammation beneath fatty tissue and other subtle structural changes in the heart muscle. “This will allow doctors to make more informed decisions about the best way to treat patients, giving the most intensive treatment to those at the highest risk,” research lead Professor Charalambos Antoniades said in the release. “We hope that, if this program is rolled out nationwide, it could reduce hospital pressures by helping patients live well for longer.”
The research team validated their tool on more than 70,000 people and is now working toward adapting it for use on any type of chest CT scan performed for any reason. “We are now working toward applying this method to any CT scan of the chest, performed for any reason,” Antoniades said.
While artificial intelligence has raised concerns about increased demand on data centers and environmental resources such as water and electricity, its potential benefits in healthcare—such as early detection of deadly conditions like heart failure—could outweigh these costs.
According to a report from Oxford University cited in their release, approximately one million people suffer from heart failure in the United Kingdom alone.
