[ad_1]

Researchers in Yale University’s Department of Cardiovascular Medicine have identified a new artificial intelligence (AI)-based video biomarker that can identify people who develop aortic valve stenosis and are at risk of rapidly worsening disease. This research was featured in a new paper, “Multi-modality video-based AI biomarkers for the onset and progression of aortic stenosis.” JAMA Cardiology April 6, 2024.

More than 1.5 million people in the United States have aortic stenosis. Aortic stenosis is a condition in which the valve that connects the heart’s main chamber to the rest of the body narrows, putting stress on the heart as it pushes blood through the body. As aortic valve stenosis progresses, pressure within the heart increases and oxygen supply to the body decreases, which can cause chest pain, difficulty breathing, light-headedness, and decreased ability to perform daily activities. Ultimately, severe aortic stenosis can lead to heart failure and early death.

Relatively new procedures such as surgery and transcatheter aortic valve replacement can effectively change the prognosis of aortic stenosis, but we do not know more about how aortic stenosis progresses in each individual. There is an unmet need to develop more accurate algorithms for proper definition.

“So far, there is no way to know who is developing aortic stenosis or who is getting worse. There are no recognized clinical biomarkers that indicate progression of aortic stenosis, and this “Most research to date has focused on repairing valves once they become diseased,” said Evangelos K. Oikonomo, MD, DPhil, clinical investigator (cardiovascular medicine) and first author. of research. “We believe this is basic research that will encourage further research into new treatments to address the progression of aortic stenosis and ultimately help prevent poor outcomes.”

Researchers built a deep learning model to detect valve features that suggest aortic valve stenosis from heart videos taken using cardiac ultrasound. Researchers have identified a signature, the Digital AS Severity Index (DASSi), that predicts which patients will progress to severe aortic stenosis.

“One of the most exciting aspects of this study is that DASSi can convert cardiac MRI, point-of-care ECG, and cardiac ultrasound all using the same model,” said Rohan Khera, MD, MSc, lead author of the paper. Associate Professor of Curriculum stated.Internal Medicine (Cardiovascular Medicine)/Director Cardiovascular Data Science (CarDS) Laboratory. “This is very important because it shows that DASSi clearly characterizes myocardial and valvular phenotypes and is not limited to modality-specific features or to some selected populations.”

The features identified by their model highlight the role of AI in detecting hidden signs of disease, rather than aspects of the image that even expert readers can visualize by watching the video. Unlike traditional methods, it doesn’t require a lot of expertise to get the views you need to get this information.

“I am extremely proud of all of our efforts at Yale to not only provide the best, least invasive care for patients with aortic stenosis, but also to continue to learn more about how the disease progresses. ” said Eric J. Velasquez. Robert W. Berliner, MD, Professor of Medicine and Chair of the Department of Cardiology at Yale University, is a co-author of this study. “This study shows that the creative use of AI can help us better understand common heart diseases and ultimately identify new approaches to reduce the number of patients who develop severe aortic stenosis. It has been shown to be useful.”

The results of this study also support the use of DASSi for opportunistic screening of aortic stenosis on ECG.

“I believe that every time a clinician undergoes a cardiac observation, it is an opportunity to screen a patient and diagnose structural heart disease,” Kela said. “This study shows that it is possible to use echocardiography and cardiac MRI to diagnose aortic stenosis and predict the risk of aortic stenosis. It has the potential to change practice. There is a gender.”

Khera, Oikonomou, and other colleagues in the CarDS lab aim to confirm the role of AI in video-based diagnosis of aortic stenosis and to identify mild or moderate cases of aortic stenosis that can progress to aortic stenosis. We plan to initiate a multi-randomized clinical trial with two goals: to identify patients with aortic stenosis. Severe aortic stenosis.

“That’s our next frontier,” Kela said. “We don’t want to stop at just building tools. We want to see if these tools and discoveries can change practice and help cardiologists better identify potential risk factors in each patient.” thinking about.”

Oikonomo recently presented his research results at the 2024 American College of Cardiology Scientific Sessions.

Other Yale researchers involved in the study include Gregory Holste. Dr. Andreas Coppi; Robert L. McNamara, MD, MHS. Dr. Norisa Haynes. Amit N. Vora, MD, MPH. Dr. Fang Lee; Thomas Gill, MD. Harlan M. Krumholtz, MD, SM.

[ad_2]

Source link