A new study has found that a special artificial intelligence (AI) tool can help predict heart attacks and death in people who have signs of stable heart disease.
The study was presented at EACVI 2025, a big medical conference held by the European Association of Cardiovascular Imaging.
Stable coronary artery disease (CAD) is a common condition where people feel chest pain from time to time, often due to narrowed heart arteries. These symptoms are usually not sudden or dangerous right away but can lead to serious problems later on.
Doctors often use a heart scan called coronary computed tomography angiography (CCTA) to check for blocked arteries. This scan gives a clear picture of the heart’s blood vessels but doesn’t always show how much the blood flow is affected.
That’s where a new AI-based tool comes in. This tool can look at the same heart scan images and estimate how well blood is flowing in the heart. The result is called FFR-CT, which stands for CT-derived fractional flow reserve. It gives doctors more information to help decide what kind of treatment a patient might need.
In the past, researchers showed that using FFR-CT helped reduce the number of extra tests that patients needed. But until now, it was not clear whether FFR-CT could also be used to predict serious health outcomes like heart attacks or death.
This new research, led by Dr. Jack Bell and his team in the UK, looked at data from a large group of patients in England. The team used records from a study called FISH&CHIPS, which included people from 27 hospitals who had FFR-CT done as part of their care.
Out of more than 90,000 people who got a heart scan, about 7,800 also had FFR-CT. Their average age was 63, and about one-third were women. Over the next three years, researchers tracked what happened to these patients.
During that time, 191 people had heart attacks, 1,573 had heart procedures like stents or bypass surgery, 74 died from heart problems, and 261 died from any cause.
The FFR-CT results were grouped into four levels: normal, borderline, reduced, and severely reduced. People with the lowest FFR-CT scores had the highest risk. For example, heart attacks happened in only 1% of people with normal FFR-CT, but in more than 5% of those with severely reduced scores.
The study showed that even people with “borderline” results had a higher risk of heart attack or death than those with normal results. This means the FFR-CT test doesn’t just help with diagnosis—it also helps doctors know which patients are at higher risk in the future.
These results stayed true even after accounting for other risk factors like age, gender, high blood pressure, diabetes, and high cholesterol. People with severely reduced FFR-CT had four times the risk of heart attack and three times the risk of dying from heart problems.
Professor Timothy Fairbairn, a senior researcher on the study, said this is the first solid proof that FFR-CT can be used to predict future heart problems. He said doctors could use this information to give extra care to people at high risk, including stronger treatment or closer monitoring.
The study also looked at whether using FFR-CT is cost-effective, and the early results suggest it could be a smart way to improve care while possibly saving money.
In conclusion, this large study shows that the AI-based FFR-CT heart scan does more than just help diagnose disease. It can also help predict who is more likely to have a heart attack or die, helping doctors make better, more personalized treatment plans.
If you care about heart health, please read studies about how eating eggs can help reduce heart disease risk, and Vitamin K2 could help reduce heart disease risk.
For more information about heart health, please see recent studies about how to remove plaques that cause heart attacks, and results showing a new way to prevent heart attacks, strokes.
Copyright © 2025 Knowridge Science Report. All rights reserved.
