What Happens When Medical Science Meets Data Science?

It’s fun to track your steps now, but wearable sensors will start to be a serious diagnostic tool.

In 2013, venture capitalist Vinod Khosla made a bold prediction: “In the next 10 years, data science and software will do more for medicine than all of the biological sciences together.”

Is that possible? Last year, wearable health sensors--products like the Jawbone Up, the Fitbit Force, and the Alivecor Heart Monitor--clocked 30 in sales and wearable technology is already used in medical research. BodyMedia’s activity-tracking armbands have been leveraged in multiple studies. The Mayo Clinic recently completed a study looking at whether data from Fitbit activity trackers can predict recovery time for cardiac surgery patients.

But medical research derived from wearable sensor data is still in its infancy. If data from personal biometric devices is ever going to be truly useful to researchers, big medical centers will have to pull it into electronic health records (EHRs), de-identify it, and make it public. Without the medical data found in EHRs, like CT scans, X-rays, and blood tests, researchers have little context for wearable sensor data and there is little useful information that can be gleaned from just the raw data--no doctor wants to deal with a patient who walks in with 15 pages about their exercise habits.

“That’s where data and predictive analytics will come in. Your Fitbit data, Withings data, data from Alivecor will probably go to the cloud, be measured, compared against millions of other people, and there will be some type of information distillation in terms of are you better or worse,” says Dr. David Albert, the founder and chief medical officer of Alivecor.

Practice Fusion, a popular EHR company, will begin opening up its API over the next year to pull in data from wearable sensors to its platform. “They may be the first organization that can get to scale with biometric data and medical record data,” says Dr. David Albert, the founder and chief medical officer of Alivecor.

Others are pushing forward as well. Basis, a startup that makes a health sensor-laden watch, is working on the first step: a device-agnostic platform that puts all of a person’s health sensor data into a single online repository. “We’re in the process of developing that for general consumption,” says Marco Della Torre, VP of Product Science at Basis. Another startup, Tictrac, lets users put data from 300 different apps, including some fitness sensor apps, into its online platform.

Albert predicts only incremental change in the wearable sensor medical research space over the next year. “Economics is what's going to drive the change. Every day, I'm told we have an unsustainable health care system in America. One of these days that light at the end of the tunnel is going to be a train, it's going to hit us, and at that point I think the motivation to utilize technology will be significantly increased. That day is coming,” he says.

[Image: Abstract via Shutterstock]

Add New Comment


  • The point at which this analysis can be used for medical research will be great! We recently did some analysis on data from wearable devices like the ones mentioned in your article and what we able to conclude was very interesting. You can see it here- http://bit.ly/1edknh3 I can only imagine how useful such analysis would be to medical providers and researchers.