Reviewed by Lexie CornerApr 18 2025
A study published in JAMA Cardiology shows that an artificial intelligence (AI) program trained to analyze images from a common medical test can identify early signs of tricuspid heart valve disease, which may help doctors diagnose and treat patients more efficiently.
David Ouyang, MD. Image Credit: Cedars-Sinai Health Sciences University
The study builds on prior research, which demonstrated that AI software can detect issues in the heart's mitral valve by analyzing ultrasound images. In this new study, researchers applied artificial intelligence to identify tricuspid regurgitation, a condition where the heart's tricuspid valve does not close fully during contraction, causing blood to flow backward and potentially resulting in heart failure.
This AI program can augment cardiologists’ evaluation of echocardiograms, images from a screening and diagnostic test that many patients with heart disease symptoms would already be getting. By applying AI to echocardiograms, we can help clinicians more easily detect the signs of heart valve disease so that patients get the care they need as soon as possible.
David Ouyang, MD, Study Senior Author and Research Scientist, Division of Artificial Intelligence in Medicine, Smidt Heart Institute, Cedars-Sinai Health Sciences University
From 2011 to 2021, researchers developed a deep-learning software designed to detect patterns of tricuspid regurgitation in 47,312 echocardiograms conducted at Cedars-Sinai.
The program identified tricuspid regurgitation in patients and classified it as mild, moderate, or severe. The software was then tested on echocardiograms it had not previously encountered, including those from patients who underwent echocardiography at Cedars-Sinai in 2022, as well as patients from Stanford Healthcare.
When compared to MRI images, the algorithm predicted the severity of tricuspid regurgitation with an accuracy comparable to that of cardiologists evaluating echocardiograms.
Future studies will focus on obtaining even more specific information about valve disease, such as the volume of blood flowing backward through a valve, and predicting outcomes if patients undergo treatment for heart valve disease.
Amey Vrudhula, MD, Study First Author and Postdoctoral Research Fellow, Cedars-Sinai Health Sciences University
Researchers at the Smidt Heart Institute are using artificial intelligence to improve a variety of cardiac imaging tests.
A major advantage of AI algorithms is that they never get fatigued and have the capacity to identify valve abnormalities from large populations of patients, taking personalized cardiology to a whole different level.
Sumeet Chugh, MD, Director, Division of Artificial Intelligence in Medicine, Smidt Heart Institute, Cedars-Sinai Health Sciences University
The study’s other Cedars-Sinai authors are Amey Vrudhula, MD; Milos Vukadinovic, BS; Alan C. Kwan, MD; Daniel Berman, MD; Robert Siegel, MD; and Susan Cheng, MD, MMSc, MPH.
Christiane Haeffele, MD, and David Liang, MD, Ph.D., are the other study authors.
The Sarnoff Cardiovascular Research Award, research grants R00 HL157421 and R01HL173526, and AstraZeneca Alexion provided funding, as well as consulting from EchoIQ, Ultromics, Pfizer, and InVision.
Journal Reference:
Vrudhula, A., et al. (2025) Automated Deep Learning Phenotyping of Tricuspid Regurgitation in Echocardiography. JAMA Cardiology. doi.org/10.1001/jamacardio.2025.0498