A new study in JAMA Cardiology shows that artificial intelligence (AI) can help find early signs of a heart problem called tricuspid valve disease. This can help doctors diagnose and treat patients faster and more accurately.
In the past, AI was used to find problems in the heart’s mitral valve by looking at ultrasound images. In this new study, researchers used AI to detect a condition called tricuspid regurgitation. This happens when the tricuspid valve in the heart doesn’t close properly, letting blood flow backward. This can lead to heart failure if not treated.
From 2011 to 2021, scientists at Cedars-Sinai trained deep-learning software to spot signs of tricuspid regurgitation in 47,312 heart ultrasound tests (called echocardiograms). The AI program was able to sort the cases into mild, moderate, or severe.
They then tested the program on new echocardiograms from 2022 and from Stanford Healthcare. When compared to MRI scans, the AI did as well as trained heart doctors in predicting how serious the valve problem was.
In the future, researchers hope to train the AI to give even more detailed information, like how much blood is leaking through the valve, and to help predict patient outcomes after treatment.
Dr. Amey Vrudhula, the study’s first author, said the goal is to improve heart care with the help of AI. Dr. Sumeet Chugh from Cedars-Sinai added that AI has a big advantage—it never gets tired and can check thousands of patients quickly, helping doctors offer more personalized heart care.
The study included several researchers from Cedars-Sinai and Stanford, and it was funded by grants and companies like AstraZeneca Alexion, EchoIQ, Ultromics, Pfizer, and InVision.
Source: azorobotics
