Severe heart valve disease is a significant cardiovascular condition requiring precise and swift diagnosis. Traditional diagnostic methods often rely on clinical assessments, medical imaging, and expert interpretation.
The emergence of artificial intelligence offers new avenues for improving the efficiency and accuracy of diagnosis. This study investigates the role of AI in automating the diagnostic process for severe heart valve disease.
Researchers at the Cardiovascular Data Science (CarDS) Lab of Yale University have found a new way to find a common heart problem called severe aortic stenosis. This problem affects the valve in the heart and can cause issues, especially in older people. Detecting it early is essential to help with symptoms and reduce the chance of going to the hospital or dying too soon.
Usually, doctors use unique heart ultrasound pictures, called Doppler echocardiography, to find this problem. However, the CarDS Lab team made an intelligent computer program using deep learning. This program can look at more straightforward heart ultrasound images and find severe aortic stenosis independently.
The person behind this technology is Rohan Khera, MD, MS. He’s a doctor who knows a lot about hearts and computers. At UT Austin, he worked with others from the Chandra Family Department of Electrical and Computer Engineering.
Researchers used many heart videos from a hospital between 2016 and 2020 to teach the computer program. Then, they tested it with more heart videos from different places in New England and California to ensure it worked well. Their findings were published in the European Heart Journal on August 23. This new way of finding severe aortic stenosis could help doctors care for their patients better.
Khera said, “Our challenge is that precise evaluation of AS is crucial for patient management and risk reduction. While specialized testing remains the gold standard, reliance on those who make it to our echocardiographic laboratories likely misses people early in their disease state.”
“We wanted to create a computer method that could be used with simple ultrasounds done right at the doctor’s office,” said Evangelos Oikonomou, MD, DPhil, who worked on the study. He’s a doctor who studies hearts and works at the CarDS Lab.
What they did helps find aortic stenosis early so patients can get help when needed. Rohan Khera, another team member, said, “Our work can help more people get checked for aortic stenosis. Handheld ultrasounds are becoming common in places like emergency rooms and clinics. You don’t always need fancy equipment.”
The most important thing about this achievement is that doctors and computer experts worked together. Greg Holste, studying for a Ph., helped a lot. He worked with Dr. Khera and made the technology even better. Dr. Khera said that “teamwork like this is essential to create new medical tools.”
The National Heart, Lung, and Blood Institute supported the study. They gave money to make this research happen and help people with heart problems.
In conclusion, the study underscores the potential of artificial intelligence in revolutionizing the diagnosis of severe heart valve disease. The developed AI algorithm showcases high accuracy and efficiency in automating the diagnostic process, which can lead to improved patient outcomes and more informed treatment decisions. As AI technology evolves, collaborations between medical experts and data scientists are crucial for harnessing its full potential in clinical practice.
Journal Reference:
- Gregory Holste, Evangelos K Oikonomou et al., Severe aortic stenosis detection by deep learning applied to echocardiography. European Heart Journal. DOI: 10.1093/eurheartj/ehad456.