Health IoT: KardiaBand sensor could replace invasive blood tests
Research appears to show that AliveCor's KardiaBand is capable of detecting hyperkalemia through a simple EKG, a feat that would previously have required an invasive blood test. 

Health IoT: KardiaBand sensor could replace invasive blood tests

Research appears to show that AliveCor’s KardiaBand technology is capable of detecting hyperkalemia – excess potassium in the blood – through a simple ECG, a feat that would previously have required an invasive blood test.

KardiaBand is a smart watch strap that provides electrocardiographic monitoring (measures electrical activity in the heart). Algorithms can then be applied to assess heart rate, stress levels, fatigue, heart age, and other health warning signs.

Read more: Health IoT: App helps sports stars predict and manage injuries

KardiaBand includes a sensor that’s compatible with the Apple Watch. Last year the company received approval from the US Food and Drug Administration (FDA) to carry out electrocardiograms using the sensor.

The test is simple: users simply place their thumb on the KardiaBand and are treated to a medical-grade ECG in 30 seconds. The band is able to detect atrial fibrillation (AF) and normal sinus rhythm.

Potential beyond ECG

However, in a presentation at the American College of Cardiology conference in Florida, AliveCor CEO Vic Gundotra unveiled new research carried out with the Mayo Clinic. The results appear to show that the same technology has potential that goes far beyond simple heart rate monitoring.

According to the research, the KardiaBand is capable of predicting dangerous levels of potassium in the blood with 90 to 94 percent accuracy. The condition, hyperkalemia, interferes with the electrical activity of cells, including those of the heart. It’s that irregular activity that the KardiaBand can pick up.

According to the research, a specific ECG pattern indicates the presence of too much potassium in the blood. Working alongside the Mayo Clinic, AliveCor used the KardiaBand to develop an algorithm that analyses users’ ECG data to spot this characteristic pattern.

Hyperkalemia doesn’t tend to cause obvious symptoms, but left untreated it can lead to kidney and heart failure. Previously, the only way to confirm its presence would have been an invasive blood test.

Read more: Flexible wearables: a game-changer for connected healthcare

Six million data-point training

To create an algorithm capable of spotting what the human eye cannot, the AI was trained using more than 23 years’ worth of medical data: two million ECGs linked to four million potassium values.

AliveCor’s FDA-cleared personal ECG technology was largely responsible for the company being named the number one Most Innovative Company in Artificial Intelligence by Fast Company last month.

“This recognition of AliveCor as the leader in artificial intelligence validates our leadership in the collection and analysis of patient-generated data, to help patients and providers assess and manage heart conditions, more conveniently, more quickly and less expensively than ever before,” said CEO Vic Gundotra.

Internet of Business says

Although not yet approved for this usage by the FDA, the KardiaBand’s ability to spot high potassium levels looks set to offer yet another compelling use case at the intersection of healthcare and consumer technology.

As a result, wearable and connected devices are increasingly able to provide both a graphical representation and data analysis point for whatever is going on inside the human body. This means that early detection, health management, and disease prevention are increasingly under the user’s control, via a variety of smart apps, which can also improve health and diet, and encourage exercise.

En masse and anonymised, this kind of data can also help societies design better, smarter, and more cost effective healthcare, if systems can be put in place to manage privacy and consent, and give something back to the data subject beyond simple advertising noise.

The development comes in a year that has already seen wearables become able to detect a variety of different medical conditions, and the development of e-skin and flexible sensor technologies.

Read more: Fitbit and Apple Watch can help predict diabetes, says report

Read more: Self-healing electronic skin brings cyborgs closer to reality