radRounds Radiology Network

Connecting Radiology | Enabling collaboration and professional development

Is AI the Future for Cardiovascular Imaging?

Bay Labs, a Bay Area startup that develops AI software to diagnose and manage heart disease, has just received Food and Drug Administration 510(k) clearance for its EchoMD AutoER product, a “fully automated clip selection of left ventricular ejection fraction.”

Ejection Fraction (EF) is the calculation of the amount of blood being pumped from the heart as it contracts. It’s the most common marker of cardiac function from echocardiograms and is a crucial component during clinical processes.

The EchoMD AutoER can be integrated into any DICOM PACS. Bay Labs co-founder and CEO, Charles Cadieu, PhD, expects the technology to help increase early detection rates and help manage patient outcomes.

The software uses deep learning to generate automated estimates of EF by assessing the entire image. The algorithm was developed through a dataset of more than 4 million images of 9,000 patients. EchoMD AutoEF digests all relevant digital video clips of a patient’s cardiac cycle from their echocardiogram study and selects the best images to measure the EF.

“Left ventricular ejection fraction has been a mainstay of echocardiography for the last 50 years. Bay Labs’ use of artificial intelligence for image selection and automated EF measurement will allow clinicians across a wide range of experience to obtain accurate evaluation of ventricular function and aid in interpretation of the echocardiograms with greater efficiency,” said Neil J. Weissman, MD, president of MedStar Health Research Institute and a professor of medicine at Georgetown University School of Medicine in a press release. “This will ultimately result in more effective care for our patients.”

Views: 143

Comment

You need to be a member of radRounds Radiology Network to add comments!

Join radRounds Radiology Network

Sponsor Ad

© 2018   Created by radRounds Radiology Network.   Powered by

Badges  |  Report an Issue  |  Terms of Service