by John R. Fischer
, Senior Reporter | September 13, 2022
Artificial intelligence proved to be better than sonographers in performing echocardiograms for diagnosing heart function.
Researchers at Cedars-Sinai Smidt Heart Institute in Los Angeles compared both types of exams in a first-of-its-kind randomized controlled clinical trial and presented the findings at the European Society of Cardiology Congress 2022 in Barcelona. They named their study, EchoNet-RCT.
They also found that cardiologists were unable to distinguish AI-based assessments from those made by sonographers.
"We asked our cardiologist over-readers to guess if they thought the tracing they had just reviewed was performed by AI or by a sonographer, and it turns out that they couldn’t tell the difference. This speaks to the strong performance of the AI algorithm as well as the seamless integration into clinical software,” said Dr. David Ouyang, a cardiologist at the Smidt Heart Institute and principal investigator of EchoNet-RCT.
The researchers looked over 3,495 transthoracic echocardiogram studies and compared assessments made with AI to exams performed by sonographers.
Cardiologists frequently agreed with the AI assessments, correcting only 16.8%, compared to 27.2% of those by sonographers. Having set out to demonstrate the non-inferiority of AI, they were “pleasantly surprised” by its superiority with respect to the prespecified outcomes, said Ouyang.
Their research was based on the use of one of the first AI technologies for assessing cardiac function. Designed by the Smidt Heart Institute and Stanford University, the solution specifically evaluated left ventricular ejection fraction, the primary measurement for determining cardiac function.
Through the current prospective and blinded study, they assessed the impact of AI when deployed in clinical settings. The scientists at Smidt Heart Institute also found that the trial could be seamlessly integrated into standard clinical workflow.
"When developed in the right way, artificial intelligence offers the opportunity to improve the quality of echocardiogram readings as well as increase efficiencies in the time and effort spent by busy cardiologists and sonographers alike,” said Dr. Susan Cheng, director of the Institute for Research on Healthy Aging at the Smidt Heart Institute and EchoNet-RCT investigator.
The Smidt Heart Institute is planning a follow-up study to evaluate how AI analysis improves clinical outcomes when applied to Cedars-Sinai echocardiogram imaging procedures.Back to HCB News