Could big data and wearables help the fight against diseases?

Biologists and medical researchers are used to drinking data with a straw. Doctors measure heart rate, weight and blood pressure, one at a time, at a high cost. When patients suffer from serious diseases, like cancer, measures are even more expensive. To make matters worse, measures are usually not shared and reused. In fact, even the patients themselves can have a hard time accessing their own data.

How do we get medical data for research? Mostly through clinical trials or one-off studies. These are excessively expensive, narrowly focused, with often very few subjects. Out of all patients suffering from serious diseases, only a small percentage will ever contribute any research data points.

Today, nearly every aspect of cancer care is based on information gleaned from the roughly 3% of patients who participate in clinical trials. But new health technologies in development offer the ability to learn from every patient. These big data tools make it possible to aggregate, analyze, and learn from a wide range of medical data—electronic health records, genetic test results, and more—while protecting the security and confidentiality of a patient’s individual data. (Masters et al., 2015)

If my car breaks down and I bring it to the garage, they can talk to the onboard computer and have much of the relevant data necessary for a diagnostic. If I were to break down in the middle of writing this blog post, the hospital would have almost no data on me.

People are more complicated than cars. Nevertheless, it seems that we are at an inflection point where much will soon become possible.

  • We have entered the era of wearable computing. Everyone is wearing a computer these days, from Syrian refugees to elderly Alzheimer’s patients. These devices range from smartphones, smartwatches, all the way to activity trackers (e.g., FitBit). We can design smart fabrics, smart glasses… All these devices are constantly connected to the Internet and have more than enough power to process the data is situ if needed.
  • The range of non-invasive medical measures that one can take continuously is expanding with every year that passes. Not long ago, just measuring your heart beat in real time required annoying straps… Yet, today, anyone with an Apple watch gets real-time heart rate monitoring. In case of cardiac problems, we can even setup people with constant 24-hour ECG monitoring if needed, and the result is reliable and practical. Google has designed glucose-tracking lenses, and they are working on cancer-tracking devices. There are apps available right now that can monitor your skin for cancer. Qualcomm has setup the Tricorder X Prize which aims to build a Star-Trek tricorder for mobile medical diagnostic. South Korean researchers have designed a brassiere that can detect breast cancer.
  • We have a fantastic cloud infrastructure that is secure and scalable. For all practical purposes, one can consider that we have access to infinite storage and computational power. If we had collected all possible data on you, and you were diagnosed with some problem, it would be a trivial matter to quickly identify similar people who had the same problem in the past, and to review the therapies that worked for them.
  • We know how to process the data in clever ways. Scientists can detect a stroke event from simple activity tracking.

So it seems that we should be entering a new era.

Conclusion: We spent the last decade with our smartest kids working in marketing for companies like Facebook. I hope that, in the next decade, they will apply their computer skills to curing the sick.

Daniel Lemire, "Could big data and wearables help the fight against diseases?," in Daniel Lemire's blog, September 28, 2015.

Published by

Daniel Lemire

A computer science professor at the University of Quebec (TELUQ).

3 thoughts on “Could big data and wearables help the fight against diseases?”

  1. Exciting and realistic vision. Yes, we can! What we need is the technical, organizational and legal ability to share the data in a responsible and safe way – we’re far from there, and it’s a necessary condotion for true medical Big data.

Leave a Reply

Your email address will not be published.

You may subscribe to this blog by email.