AppleWatch wristband sensor claims to detect potassium in your blood — without needles

Bose’s new augmented reality glasses use sound instead of sight
March 12, 2018
Immersive experience: I found a stranger from another dimension in my hotel room shower
March 12, 2018

AliveCor KardiaBand, a sensor compatible with Apple Watch, can detect dangerous levels of potassium in the blood with 94 percent accuracy. Although the US Food and Drug Administration UU KardiaBand has not yet been approved for this purpose, it is an interesting step forward since, at this time, the condition is usually detected by invasive blood tests using needles.
The AliveCor KardiaBand is a sensor that is inserted into a slot in the watch strap. The user touches the sensor, which then takes a reading of the electrical activity of the heart, called electrocardiogram (EKG). This reading can reveal an abnormal heart rhythm and atrial fibrillation (AFib), and the sensor sends the information to an application. Yesterday, at the conference of the American College of Cardiology in Florida, the CEO of AliveCor, Vic Gundotra, presented an investigation carried out with the Mayo Clinic that shows that the same technology can detect too high levels of potassium in the blood, called hyperkalemia.
Hyperkalemia can be caused, among other things, by diabetes, dehydration and chronic kidney disease. It can cause kidney and heart failure and does not cause obvious symptoms, which means you may have the condition and not know it.
Too much potassium interferes with the electrical activity of cells, including cardiac cells. This means that it is dangerous to the heart, but it also means that high levels of potassium change the electrical reading of the heart, which means that a certain ECG pattern can reveal the presence of too much potassium, according to Gundotra. AliveCor worked with the Mayo Clinic to develop a new algorithm for KardiaBand that can analyze ECG data and detect if the user has hyperkalemia. The data set included 2 million EKGs linked to 4 million potassium values, which were collected over 23 years.
To train AI with these data points, the team took the data set and divided it into parts. They used some of the data to train the network. Basically, they indicated which patterns of EKG reading showed hyperkalemia, and let the AI ​​learn by itself how to detect the pattern. Once the training was completed, the team tested the AI ​​on a different part of the data to see if, given only the ECG, they could tell if it revealed hyperkalemia. It had an accuracy of between 90 and 94 percent.
Some previous research has suggested that electrocardiograms may not be a good way to diagnose hyperkalemia, but, to be fair, that research was very limited and tested two human doctors. Another study suggested that EKG readings may not be sensitive enough to trap everyone with hyperkalemia and that the condition does not always cause a different ECG reading.
It will be a while before we see this new technology become common. Last November, the FDA eliminated the KardiaBand as the first medical device that works with the Apple Watch, but Gundotra emphasizes that the results do not mean that the KardiaBand has already been approved by the FDA for the diagnosis of hyperkalemia. They will work on that and build more clinical trials.


Leave a Reply

Your email address will not be published. Required fields are marked *

I accept that my given data and my IP address is sent to a server in the USA only for the purpose of spam prevention through the Akismet program.More information on Akismet and GDPR.