Over the weekend AliveCor announced that its technology was validated in two studies. The first found that its KardiaBand, which works as a personal electrocardiogram device, can accurately detect atrial fibrillation (AFib) when used with the Apple Watch. The second study, which was a collaboration with the Mayo Clinic found that its ECG device, when paired with AI, could detect hyperkalemia.
“As our team continues to push the bounds on innovation in digital health, we are on a path to changing the way AFib and hyperkalemia can be detected, and to defining the ways in which products like Apple Watch can play a role in the future of health care,” Vic Gundotora, CEO of AliveCor, said in a statement.
Atrial Fibrillation study
The Cleveland Clinic study, which has been accepted for publication in the Journal of the American College of Cardiology, found that the KardiaBand successfully detected atrial fibrillation and normal sinus rhythm with an accuracy level that was comparable to physicians interpreting the same ECGs, according to a statement. In fact, researchers found that the Kardia algorithm correctly interpreted atrial fibrillation vs normal sinus rhythm with 93 percent sensitivity and 84 percent specificity.
The band, which was first launched in November of 2017, is the first FDA-cleared medical device for the Apple Watch. Users swap the KardiaBand for the traditional Apple Watch band.
But this isn’t the only system looking to detect atrial fibrillation. In May Cardiogram, a startup working on algorithms to make the Apple Watch’s heart rate data clinically actionable, announced results from its mRhythm Study, which showed the company’s algorithms can detect atrial fibrillation with 97 percent accuracy. At the time researches said they would continue to hone the Cardiogram algorithm and explore the potential of using a deep neural network to detect other conditions.
Apple has become increasingly involved in atrial fibrillation detecting technologies. Notably In November the tech giant announced the Apple Heart Study, an Apple Watch-based ResearchKit study that uses the heart rate sensor to look at potential arrhythmias. The study, which is opened to anyone in the US who has an Apple Watch Series 1 or later and is over 21, continuously monitors participants heart rates. If the Apple Watch detects possible atrial fibrillation, the user will get a notification on their Apple Watch or iPhone, which will prompt them to connect with a doctor via telemedicine.
Mayo Clinic uses technology to research potassium levels
In addition to the Cleveland Clinic study, the company announced a second study by the Mayo Clinic, which demonstrated that when paired with artificial intelligence AliveCor’s ECG device is able to detect hyperkalemia, where individuals have high potassium levels in their blood.
“The Mayo Clinic partnered with AliveCor...on trying to do something extraordinary, which was to to look at the ECG use artificial intelligence and hope the AI would be able to look at a single ECG and be able to tell you your potassium level,” Gundotra told MobiHealthNews. “That is kind of crazy because [calculate] potassium levels you usually have a blood test. High potassium or low potassium can be life threatening."
Researchers found the sensitivity for hyperkalemia detection was between 90 and 94 percent.
In the study the artificial intelligence algorithm that was used to detect hyperkalemia was developed by taking data from at 2 million ECGs linked to 4 million serum potassium values collected over 13 years, as well as prospective data from an AliveCor smartphone device, according to the company. AliveCor claims that up until the KardiaBand was developed, the only way to test for hyperkalemia was through blood testing.
“This is totally new,” Gundotra said. “Two years ago people would have thought this was science fiction and until we get FDA clearance it’s not real. But I think the first step is publishing your results and this is a major milestone.”
Results from both of the latest KardiaBand studies will be presented at the American College of Cardiology’s 67th Annual Scientific Session. While both studies are promising for AliveCor, Gundotra said he is taking it in stride. There is a long road to go, especially for the AI detecting potassium levels, which isn't yet FDA-cleared.
“This is medicine, this is people’s lives. There is a thoughtful process, you publish your data and have to work with trust partners and work with the FDA,” Gundotra told MobiHealthNews. “AliveCor trusts the process. Sometimes takes a little longer than we would like but you are dealing with human lives and so we appreciate why it takes that long.”