Dutch researchers use AI to predict repeat heart attacks, potentially saving lives
Researchers at Radboud University Medical Center (Radboudumc) have developed an artificial intelligence (AI) system that can predict which patients are at risk of a second heart attack, requiring another angioplasty, or dying within two years after an initial heart attack, according to AD.
In the Netherlands, about 40,000 people undergo angioplasty each year after blood clots in coronary arteries cause heart attacks, restricting oxygen to the heart.
During the procedure, arteries are widened with a balloon and often a stent is placed. Yet 15 percent of patients experience another heart attack within two years, sometimes fatal, partly due to weak spots in artery walls.
Radboudumc researchers used images from an optical coherence tomography (OCT) camera, a miniature device threaded through arteries via the wrist, to detect these vulnerable areas.
“It can produce more than 700 images, far too many for a human to assess quickly during surgery,” Jos Thannhauser, a biomedical engineer involved in the study, told AD.
Working with physician-researcher Rick Volleberg, Thannhauser applied AI to review the images, monitoring more than 400 patients in over 2 years. The study found that AI predicted new heart attacks more accurately than specialized labs that normally analyze OCT scans.
“Now that we know this, we can in the future intensify medication, target lifestyle changes, or even place preventive stents for high-risk patients. That can truly save lives,” Thannhauser said.
