Why Supercomputer Watson Wants Your Medical Records

Clockready, Wikimedia Commons // CC BY-SA 3.0
Clockready, Wikimedia Commons // CC BY-SA 3.0 / Clockready, Wikimedia Commons // CC BY-SA 3.0

It's been a few years since IBM’s Watson computer crushed the competition on Jeopardy!, and since then, its cognitive computing powers have been redirected to the field of medical science.

In 2013, doctors and nurses at Memorial Sloan-Kettering Cancer Center started using Watson to help determine the best course of care for cancer patients. The computer “learned” through scanning things like textbooks, scientific journals, and medical histories, and then used that information to generate treatment ideas. Doctors weighed that analysis with their own and made decisions on how to proceed.

Now, in conjunction with CVS Health, Watson will study the medical history of people with chronic conditions like hypertension, heart disease, diabetes, and obesity to find patients who seem most likely to need emergency healthcare in the future. In doing so, they hope to prevent such an outcome by recommending preventative care to the right people.

CVS will tap into the Watson Health Cloud—a data resource with de-identified patient information to keep the privacy and security of individuals protected. It's HIPAA-compliant, and the entire system will work on an opt-in basis, so existing patients don't have to worry about their information getting scanned without their express consent.

The timeline for the system is loose, but Troyen A. Brennan, chief medical officer for CVS Health, said in an interview with The Washington Post that “realistic interventions” could be identified in a year or two.

Eventually, the system could expand to include things like a mobile app that connects to a patient's fitness trackers, regularly adding data to create an even fuller picture of health.

In a news release, senior vice president of IBM Watson Mike Rhodin said, “The capabilities of the IBM Watson Health Cloud, when coupled with CVS Health’s insights into medication utilization and patient behavior, could prove transformative for the industry.”