FDA clears AI model for detecting signs of heart failure in ECGs

Anumana, a Massachusetts-based nference portfolio company, has received U.S. Food and Drug Administration (FDA) clearance for its new artificial intelligence (AI) model that identifies signs of low ejection fraction (LEF) in 12-lead electrocardiograms (ECGs).

The ECG-AI LEF algorithm, developed as part of a collaboration with Mayo Clinic in Rochester, Minnesota, was designed to help screen high-risk patients for heart failure. It was built using data from more than 100,000 ECGs and 100,000 echocardiograms and has already been tested in a clinical setting on more than 40,000 patients. The AI model was then validated in a clinical trial involving 16,000 patients, producing a sensitivity of 84.5% and specificity of 83.6% when it came to spotting patients with an ejection fraction less than or equal to 40%.

“Anumana’s ECG-AI LEF fills an important unmet need—the lack of an easily accessible point-of-care, noninvasive, and inexpensive tool to screen for a weak heart pump,” Paul Friedman, MD, chair of the department of cardiovascular medicine at Mayo Clinic and chair of the Anumana Board of Advisors, said in a prepared statement. “It allows identification of otherwise hidden disease, for which many effective, lifesaving treatments are available—once the presence of the disease is known.”