The focus on wearable health tech started with the hardware before shifting to the software, as is often the case with new technology. With personal computers in the early to mid 80s, for example, the key question for most people was what brand machine to buy, not which software would do the job. The same arc is already apparent with wearables; brands and form factors get the early headlines. However, as our editor Alfred Poor asserts regularly, the value is in the algorithms. The biometric sensors in wearables gather and transmit data, but the algorithms in the associated software do the heavy lifting. Algorithms increasingly call on artificial intelligence in the form of deep learning and neural networks to pull usable information from masses of data.
A new Android smartwatch app demonstrates the relative roles of wearable health tech hardware, algorithms, and artificial intelligence. Cardiogram, Inc. just announced Cardiogram for Android Wear. Previously, Cardiogram was only available for Apple Watches. Cardiogram’s mRhythm project, which we covered earlier this year, used a deep learning algorithm called DeepHeart to detect atrial fibrillation (AFib). Undiagnosed AFib is a leading cause of strokes. In clinical tests published by the Heart Rhythm Society, DeepHeart detected AFib with 97% accuracy using the Cardiogram app with Apple Watches.
Compatible with any Android Wear watch with a heart rate sensor, Cardiogram for Android gives users insights into health including sleep, fitness, stress, and medical conditions. According to the developers, Apple Cardiogram app users told the company they had used it to monitor PTSD, diabetes, sleep apnea, and other medical conditions. The app default is to take background heartrate measurements every five minutes. Cardiogram for Android users can adjust the time interval up or down to their preference or as desired for specific purposes.
