AI Transforms Smartwatch ECG Signals Into A Diagnostic Tool For Heart Failure

A study published in Nature Medicine reports on the ability of ECGs from "smart" watches to accurately identify heart failure in non-clinical settings. Mayo Clinic researchers applied artificial intelligence (AI) to Apple Watch ECG recordings to identify patients with weak heartbeats. Study participants record the EKG of their smartwatch remotely whenever and wherever they want. They periodically upload EKGs automatically and securely to their electronic medical record through a smartphone app developed by the Mayo Clinic Center for Digital Health.

"Currently, we diagnose ventricular dysfunction - a weak heart pump - with an echocardiogram, CT or MRI, but these are expensive, time-consuming and sometimes unavailable. The ability to remotely diagnose a weak heart pump with an ECG. Recorded by a person with a consumer device such as a smart watch the information gained allows early detection of this potentially life-threatening disease on a large scale," said Paul Friedman, M.D., chief of cardiovascular medicine at the Mayo Clinic in Rochester and lead author of the study.

People with a weak heart pump may have no symptoms, but this common form of heart disease affects about 2% of the population and 9% of people over the age of 60. When the heart cannot pump enough oxygen-rich blood, symptoms such as shortness of breath can develop. breathing, palpitations and swelling of the legs. Early diagnosis is important because once detected, many treatments are available to improve quality of life and reduce the risk of heart failure and death.

Mayo researchers interpreted the Apple Watch's single-lead ECG by modifying an earlier algorithm developed for 12-lead ECGs that had been shown to detect a weak heart pump. The 12-Lead Low Ventricular Ejection Fraction Algorithm is licensed from Anumana Inc., an AI-powered health technology company founded by nference and the Mayo Clinic.

Although the data is still preliminary, a modified AI algorithm using single-lead ECG data has an area under the curve of 0.88 to detect a weak heart pump. In comparison, this accuracy score is equal to or slightly better than the treadmill medical diagnostic test.

"These data are encouraging because they show that digital tools allow for convenient, inexpensive and scalable screening of critical conditions. Technology allows us to collect useful information about a patient's heart remotely in an accessible way that meets the needs of people everywhere. Principal Investigator of the Division of Cardiovascular Medicine at the Mayo Clinic, Ph.D. Zachi Attia says. Dr. Attia is the first author of the study.

“As demonstrated in this study, the ability to remotely extract data from portable consumer electronic devices and provide analytics to prevent disease or improve health has the potential to revolutionize healthcare. Solutions like these not only help predict and prevent problems, but ultimately help reduce health inequalities and the burden on healthcare systems and doctors," he said. learn. .

All 2,454 study participants were Mayo Clinic patients from the United States and 11 countries. They downloaded an app developed by the Mayo Clinic Center for Digital Health to securely upload EKGs from their Apple Watch to their electronic medical records. From August 2021 to February 2022, participants entered more than 125,000 old and new Apple Watch ECGs into their electronic medical records. Physicians can access all ECG data, including date and time of collection, in an AI dashboard embedded in the EMR. is registered.

About 420 participants completed an echocardiogram, a standardized test that uses sound waves to visualize the heart, within 30 days of recording their Apple Watch ECG in the app. Of these, 16 patients had low ejection fractions confirmed by echocardiogram, providing comparable accuracy.

This research was funded by the Mayo Clinic with no technical or financial support from Apple. Doctors Attia and Friedman, among others, are co-inventors of the low ejection fraction algorithm licensed by Anumana and may benefit from its commercialization.

story source:

Materials provided by Mayo Clinic . Originally written by Terry Malloy. Notes. Content is subject to change in style and length.