If you’ve had an electrocardiogram (ECG) in a clinical setting, a technician most likely attached 10 electrodes to your chest and limbs. That array of electrodes is the standard setup for a 12-lead ECG. The numbers are different which can be confusing, but in this sense a “lead” is synonymous with a “view” instead of an attachment (the electrodes). A cardiologist reads a 12-lead ECG for signs of many conditions including arrhythmias and acute myocardial ischemia (limited blood flow). ECGs also yield other information about heart health from the organ’s electrical activity. So, if the current standard of care for assessing heart health calls for 10 electrodes attached to your bare skin, each in carefully determined locations, how can smartwatches and smart patches that tout single-lead ECG capabilities be all that useful? Mayo Clinic researchers recently demonstrated that significantly useful heart health data is indeed detectable with a single-lead ECG.
On May 1, Mayo Clinic researchers presented a study abstract at Heart Rhythm Society conference. The abstract summarized early results of a study using a modified version of a recognized 12-lead artificial intelligence algorithm that can detect low ventricular ejection fraction: a weak heart pump in simple terms. The algorithm licensee is Anumana Inc.; Mayo Clinic personnel modified the algorithm to work with single-lead data from an Apple Watch.
Itzhak Zachi Attia, PhD is the co-director of Artificial Intelligence in the Department of Cardiovascular Medicine at Mayo Clinic. Dr. Attia modified the 12-lead algorithm by creating a technique that adapts the Apple Watch’s single-lead data into separate signals the standard algorithm can read. Mayo Clinic Center for Digital Health personnel developed a smartphone app to capture Apple Watch ECG data and transmit it to a Mayo Clinic data platform. Researchers analyzed data from 2,454 Mayo Clinic patients who had an iPhone, the Mayo Clinic app, and an Apple Watch Series 4 or later. The team also used data collected previously during a six-month study that yielded 125,610 ECGs from 46 states and 11 countries including 420 patients who had a watch ECG within 30 days of an echocardiogram. Echocardiograms measure heart pump strength.
The Mayo Clinic researchers state that while the data analysis is still in preliminary form, the Apple Watch ECG data correlates with the echocardiogram results at a rate of 0.88. This high performance level is similar or a bit better than a medical treadmill test for identifying a weak heart pump, according to Mayo Clinic researchers.
These are early days in using clinically-significant sensor data from relatively inexpensive consumer wearables. However, but as someone with low ventricular ejection fraction, in the past 12 months I have had two CT scans, two echocardiograms, multiple X-rays, and a nuclear stress test. I wore a continuous heart rate monitoring Zio patch for two weeks. The chance that the time and hassle of all that testing might be replaced by a smart watch with a single-lead ECG, a smartphone, and a mobile app is a jaw-dropper. Count me in.
