Study shows AI can “extend the reach” of echocardiographic assessment

Caption AI software in action showing a successful ultrasound.

Caption Health has announced the publication of a study in JAMA Cardiology assessing its artificial intelligence (AI) guidance software for cardiac ultrasound examinations. The study was the basis for the February 2020 authorisation of Caption Guidance through the US Food and Drug Administration’s (FDA) de novo pathway. The Caption AI platform, which includes Caption Guidance and Caption Interpretation, is an AI-guided medical imaging acquisition software.

The study demonstrates that ultrasound images captured by nurses without prior ultrasound experience and reviewed by experienced cardiologists were shown to be of diagnostic quality to assess left ventricular size and function in 98.8% of patients, right ventricular size and function in 92.5% of patients, and in 98.8% of patients for presence of pericardial effusion. The study was conducted with 240 patients aged 20‒91, 42% female patients, with 17.6% of patients Black or African-American, and 33% of patients with a BMI of 30 or greater.

“In our mission to democratise access to healthcare and quality medical imaging, we wanted to ensure that we tested Caption Guidance on a wide range of patients to prove its effectiveness across a diverse population,” said Yngvil Thomas, head of Medical Affairs & Clinical Development at Caption Health. “This study shows that AI-guided imaging can expand healthcare professionals’ skill sets in a meaningful way with minimal training—giving patients more opportunities to receive timely diagnostic care.”

Each patient in the study underwent paired ultrasounds: one from a nurse, and one from an experienced registered diagnostic cardiac sonographer. In addition to evaluating diagnostic quality, the cardiologists also made diagnostic assessments. For the diagnostic assessments corresponding to the primary endpoints listed above, there was at least 92.5% agreement between the nurse and sonographer scans.

The results indicated that Caption Guidance performed well for patients with various cardiac pathologies that might be encountered in real-world clinical practice; more than 90% of patients were found to have cardiac abnormalities in scheduled full echocardiograms performed within two weeks of the study. The results were also consistent across BMI, sex, and race, further demonstrating Caption Guidance’s efficacy and robustness.

“The study’s remarkable agreement between nurses’ scans and sonographers’ scans shows that the use of AI like Caption Guidance could fundamentally change how we use medical imaging,” said Akhil Narang, a cardiologist at Northwestern Medicine, Chicago, USA and first author on the paper. “This will extend the abilities of healthcare providers to evaluate for different pathologies in critical care, emergency departments and other settings—and perhaps identify them even earlier with the assistance of AI.”

Given the demonstrated effects of COVID-19 on the heart, the study notes that technology like Caption Guidance may allow for reduced exposure of sonographers to the disease. Caption AI has been used in hospital systems for assessment and treatment of the cardiovascular impact of COVID-19, and is quickly becoming the new standard of care in critical care settings for timely diagnosis and management.

“Cardiac ultrasound is a powerful diagnostic technology that can be helpful, and even life-saving, in many clinical settings,” said  Chris Moore, an emergency medicine physician at Yale School of Medicine, New Haven, USA. “However, its application may be limited by the availability of experienced users. This groundbreaking study represents a significant step forward in making this technology more widely accessible to patients wherever needed.”

In addition to further refinements of its cardiac ultrasound guidance, Caption Health recently received a US$4.95 million grant from the Bill & Melinda Gates Foundation to develop AI-guided lung ultrasound software.


Please enter your comment!
Please enter your name here