Session

Continuous Digital Biomarkers from Wearable Devices

Brandon Ballinger · Cardiogram

We’re all familiar with getting our blood drawn at the doctor’s office. What if wearables could tell us our digital biomarkers continuously and painlessly?

We’ll report on a collaboration between Cardiogram and UCSF Cardiology, touching on deep learning, wearable heart rate sensors, and healthcare. We’ll describe some well-understood biomarkers from the medical literature—resting heart rate, heart rate variability, and heart rate recovery—and first show how these translate into consumer-grade wearables such as Apple Watch and Android Wear.

Then, we’ll dive into applications across the spectrum of health: from how continuous digital biomarkers can predict disease states, to how activities like meditation or exercise can lift you up to the top percentile. We’ll wrap up with some lessons learned bridging the worlds of machine learning and medicine.