AI flags patients at increased risk of CVD
Researchers have trained a machine learning model to identify patients with familial hypercholesterolemia (FH), a genetic disorder that increases a person’s risk of coronary artery disease.
The study, published in The Lancet Digital Health, was funded by the FH Foundation, a research and advocacy group focused on FH.
“Timely identification of FH and initiation of guideline-based therapies can markedly attenuate the risk of coronary artery disease,” wrote lead author Kelly Myers, chief technology office of the FH Foundation, and colleagues. “Unfortunately, fewer than 10% of individuals with FH have been identified, creating a huge reservoir of unidentified and untreated individuals with the condition.”
The group’s machine learning model was trained on data from more than 900 patients with FH and more than 83,000 patients “presumed free of FH” from four different institutions in the United States. It was then applied to one national database of 170 million patients and a separate dataset of 174,000 patients from the Oregon Health & Science University electronic health record database. All patients used for this study had “at least one cardiovascular disease risk factor.”
Specialists reviewed all cases flagged by the machine learning model, noting that it achieved an accuracy of 87% for the national database and an accuracy of 77% for the Oregon Health & Science University database.
“Precision screening for FH is now a reality in any healthcare system with electronic health records,” co-author Daniel J. Rader, MD, chair of the department of genetics in the Perelman School of Medicine at the University of Pennsylvania and chief scientific advisor of the FH Foundation, said in a prepared statement. “We no longer need to screen everyone to find individuals who are at genetic risk for heart attacks and strokes. After further clinical evaluation, if an FH diagnosis is made, it will trigger screening of relatives as well. While FH is manageable, the greatest benefit is from treatment earlier in life.”
Michael has more than 18 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.