AI models can guide nurses with no prior experience through the process of obtaining 10-view transthoracic echocardiographic (TTE) studies, according to new research published in JAMA Cardiology.
“In many clinical settings, echocardiography is unavailable because of a lack of trained personnel,” wrote lead author Akhil Narang, MD, of the Bluhm Cardiovascular Institute at Northwestern University in Chicago, and colleagues. “In these settings, nonexpert users may perform limited examinations using handheld or portable machines, but quality is nonuniform, with risks for nondiagnostic and misleading imaging.”
To see if AI could improve that ongoing dilemma, Narang et al. turned to software that was developed to “emulate sonographer expertise” by monitoring image quality and providing helpful cues when necessary.
The team trained its AI model on images from multiple vendors, making it usable with a variety of platforms, and tested its effectiveness on a group of 8 nurses with no previous experience conducting echocardiograms. Each nurse scanned 30 adult patients who had been scheduled to undergo an echocardiogram at one of two facilities from March to May 2019.
A team of five experienced specialists independently reviewed each nurse’s acquisitions. Overall, 98.8% of scans were of diagnostic quality for left ventricular size, function and pericardial effusion and 92.5% were of diagnostic quality for right ventricular size.
“This study demonstrates that an AI algorithm can guide novices, without prior ultrasonography experience, to acquire 10 standard TTE images that, when analyzed together, provide limited diagnostic performance for the evaluation of left ventricular size and function, right ventricular size, and presence of a nontrivial pericardial effusion,” the authors wrote.
Narang and colleagues did emphasize that their deep learning model “was developed to extend echocardiography access, but not to replace sonographers who provide expert imaging.” In addition to improving access, the researchers also think their work could potentially help users provide better patient care. If the AI model can help nurses with no experience at all, it might even just as helpful for users who already have some experience, providing them with important reminders that they can carry with them moving forward.
The full analysis is available here.
Michael has more than 16 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.