AI combined with ultrasound could help diagnose cardiomyopathies earlier. [Photo by Peakstock via Stock.Adobe.com]
The experimental test could ensure more people receive care for two types of cardiomyopathy. Investigators published findings from a study of the technology in The Lancet Digital Health.
Researchers at the Cardiovascular Data Science (CarDS) Lab at YSM say they developed an algorithm to recognize signs of two common types of often underdiagnosed cardiomyopathies in ultrasound imaging captured during emergency room visits over the last decade across two health systems. AI picked up signs of disease an average of two years before the patients were diagnosed.
“AI-based tests like this one could be deployed in emergency rooms to flag people at high risk of cardiomyopathies,” said Dr. Evangelos Oikonomou, a clinical fellow in cardiovascular medicine and postdoctoral fellow in the CarDS Lab at YSM. “We can actually minimize how many cases fall through the cracks.”
According to YSM, ultrasounds for heart conditions usually last around 30 minutes. But, those who visit emergency rooms sometimes receive cheaper, faster ultrasounds of their chests. These aren’t meant to pick up cardiomyopathy, YSM says, but could offer an opportunity to screen for cardiomyopathy risk.
To test their hypothesis, the team focused on hypertrophic cardiomyopathy (where heart muscles thicken to the point that the heart has trouble beating) and transthyretin amyloid cardiomyopathy (where a protein buildup causes the heart muscles to stiffen).
Oikonomou and the CarDS lab developed their algorithm using more than 90,000 brief ultrasounds collected over a decade. Around 550 people in that dataset later received a diagnosis for one of the two cardiomyopathies. After training, the algorithm correctly flagged almost all positive cases. Screening also picked up early signs of illness in several ultrasounds captured between six months and four-and-a-half years before the patients received an official diagnosis.
YSM says this suggests that the algorithm may spot things experts fail to notice. This could prove crucial — the team says early intervention can increase survival odds for transthyretin amyloid cardiomyopathy by 30%.
The team ultimately hopes to develop a cheap, easy way to screen for cardiomyopathies while performing other care. Once flagged, people can go for further testing to confirm their condition. AI won’t replace diagnosis or medical treatment but can enable patients to access the care they need sooner.
“Why do AI?” Oikonomou said. “Because it can pick up things from these images that human experts—even us cardiologists—cannot.”
