Data and Heart Disease: AI in Cardiovascular Medicine
Summary
- Dr. Yiannis Chatzizisis is leading a Miller School Division of Cardiovascular Medicine that is using AI to solve heart care problems.
- Dr. Chatzizisis recently moderated a series of technology-based discussions at the American Heart Association Scientific Sessions.
- Dr. Chatzizisis says technology is democratizing heart care and putting specialized knowledge in the hands of more providers.
Imagine predicting a heart attack being as straightforward as GPS turn-by-turn guidance and life-saving procedures taught without scalpel touching the skin.
That future is unfolding for the University of Miami Miller School of Medicine’s Division of Cardiovascular Medicine, and professor and chief Yiannis S. Chatzizisis, M.D., Ph.D., is leading the charge.
Dr. Chatzizisis is an internationally recognized leader in the use of artificial intelligence (AI) in cardiovascular care. His research has garnered more than 12,000 citations and he regularly leads panels on AI interventions in cardiology at professional conferences and forums worldwide, including the most recent American Heart Association Scientific Sessions.
“I moderated about 10 presentations that included esteemed institutions from around the country, the majority of which discussed the application of AI in cardiovascular medicine,” Dr. Chatzizisis said. “These are applications, technologies, tools that can help us better diagnose coronary artery disease, particularly heart attacks.”
Using EKG Data to Learn About Heart Risk
He sees great potential in source material cardiologists already have in abundance.
“We can use deep learning models to extract signals that are hidden in EKG data to learn about risks of heart attack and occlusive coronary artery disease,” Dr. Chatzizisis said.
AI-powered tools can interpret complex heart images from both non-invasive scans (such as echocardiograms and CT coronary angiograms) and invasive methods (such as intravascular ultrasound). By merging images from different techniques, AI helps create precise, 3D models (so-called digital twins) of heart vessels and can identify blockages and plaques, which could cause heart attacks.
What’s at stake with such a diagnostic ability is nothing less than life and death. Dr. Chatzizisis uses a common scenario to amplify the importance of deploying the full strength of AI on behalf of patients.
“We can use technology to predict whether a heart attack is going be severe, one that requires immediate intervention,“ Dr. Chatzizisis said. “If you’re in the field, a paramedic, and you know how to run an application that tells you this information…we’re talking about saving the patient’s heart, and life. Leveraging big data is, often, repurposing existing information for better patient care.”
Extending Knowledge About Heart Care
The importance of the information AI provides is underscored by a democratization of knowledge formerly reserved for people with the training and experience of Dr. Chatzizisis. Reserved, that is, for the small sliver of humanity that can call themselves an unqualified expert. AI-powered interfaces can give a non-expert the same information about a cardiac computed tomography test that imaging experts have acquired through years of training.
“It’s a revolution,” Dr. Chatzizisis said.
But a revolution with a few caveats, he warned. AI is only as effective as its source data allows. Health care must be cognizant of the data sets it feeds to AI-powered applications that guide its work.
“If you use big data from the U.S., you’re probably working just with the U.S. population,” Dr. Chatzizisis said. “But to translate this to a Chinese population is a logical leap that may not work.”
And effective big data manipulation is costly, at least now. But Dr. Chatzizisis believes the natural curve of technology will reduce expense over time.
“It’s like the iPhone,” he said. “Over time, they became more cost-effective.”
AI in Patient Treatment
With ingenuity and resolve, barriers dissolve. And, according to Dr. Chatzizisis, the potential of AI-guided cardiovascular care will go well beyond interpreting heart images and stress tests.
“The real question is, ‘How can AI be integrated into treatment?’” he said.
It’s a question he and Miller School colleagues at the Center for Digital Cardiovascular Innovations are beginning to answer. Dr. Chatzizisis already uses big data to refine stenting for bifurcated arteries by creating the aforementioned digital twin. The computer model of an individual’s heart allows virtual testing and use of tailored stents to predict how different stent designs and techniques will perform.
Under Dr. Chatzizisis’ leadership, the center uses technology to help patients who need coronary artery bypass surgery. Virtual models assist in helping predict how different bypass grafts will function by analyzing blood flow and artery compatibility. This can help shorten surgery time, make the procedure safer and improve patient recovery.
The AI horizon is infinite, and Dr. Chatzizisis is part of both a local Miami team and international cadre of the field’s finest minds whose charge it is to transform infinite potential into concrete reality.
“This is another tool, another ally, another weapon in our hands that helps us offer better care to our patients,” he said.
Tags: AI, artificial intelligence, Center for Digital Cardiovascular Innovations, Division of Cardiovascular Medicine, Dr. Yiannis Chatzizisis, heart, heart attack, technology