Artificial intelligence diagnostics company Cardiologs today announced the results of a clinical study showing its AI-based ECG analysis software “significantly” reduces the rate of false-positive atrial fibrillation (AFib) detection in remote cardiac monitoring of certain patients.

The company, which has offices in Paris and Boston, worked with Valley Health System of northern New Jersey to track a cohort of 425 patients who each received a Medtronic Reveal LINQ implantable loop recorder (ILR) for known AFib or cryptogenic stroke.

Researchers evaluated 1,500 AF episodes and annotated each as either true or false, according to the study, published in the Journal of the American College of Cardiology Clinical Electrophysiology. All applicable ECG readings were uploaded into Cardiologs’ cloud-based solution for analysis by its deep-learning algorithm. The research showed Cardiologs’ AI model eliminated the incidence of false-positives by as much as 66% and had a positive predictive value (PPV) as high as 75%.

According to the American Heart Association, more than 2.7 million Americans are living with AFib, and that number continues to grow. ILRs play a critical role in the management of these patients. Devices with wireless capabilities are programmed to automatically transmit ECG data to a secure website for review by cardiology staff. However, the large volume of recordings can present challenges — in particular, the substantial number of false-positives alerts that electrophysiologists must review to avoid misdiagnosis and potential errors in clinical management.

The high false-positive rate of AFib detected by ILRs has created a significant clinical burden, according to lead investigator Dr. Suneet Mittal, director of electrophysiology and the Snyder Center for Comprehensive Atrial Fibrillation at Valley Health System.

“Since ILRs transmit data daily, these false positives are one of the Achilles heels of remote cardiac monitoring,” Mittal added in a news release. “This study validates that Cardiologs’ advanced AI can filter 2/3 of false-positive AF episodes, which should improve clinical efficiency.”

“The need for reliable and efficient remote patient management is stronger than ever before and will only increase in the future,” added Cardiologs CEO Yann Fleureau. “Publication of our successful trial results in JACC EP further validates the considerable promise of our solution for the clinical community. We have demonstrated that we can offer an extremely reliable solution that will reduce the clinical burden of managing ILR patients remotely.”