Researchers at the University of Oxford may have just found a new way to use AI to our advantage.
Heart attack “red flags” have been successfully identified by a machine that creates a fingerprint called a “fat radiomic profile.” The red flags that it identifies range from inflammation and scarring to changes in blood vessels that are used to supply blood to the heart. All of these, as you would imagine, also indicate the potential for a future heart attack.
Typically when patients go to the hospital experiencing pains that could mimic a heart attack, their coronary arteries are scanned through a coronary CT angiogram. If nothing shows up on that scan, patients are typically sent home. Unfortunately, for many of those patients, hidden heart attack symptoms still exist.
Researchers hope that by using fingerprint artificial intelligence, doctors will be able to find any of the bad characteristics that could exist in people’s arteries. 85% of deaths are related to heart attacks and strokes. Being able to catch problems early will have a profound effect on the health of people.
The results from the scans could refine the accuracy of diagnoses as well as provide important information at an early stage of the disease. This would give cardiologists the ability to predict what could happen in the future and start treatment strategies to potentially save people’s lives. This AI could be just what the doctor ordered.