IBM Watson Health Tackles Diabetes

IBM Watson: Machine-Of-All-Trades

Where do wearables, healthcare, and big data meet? One place is in research and care for patients suffering from diabetes, which is considered one of the most common and costly chronic conditions today. In the US, 1.4 million people are diagnosed with diabetes each year, according to the American Diabetes Association (ADA).

In 2012, according to the ADA, 29 million people were living with the disease, and another 86 million were diagnosed with a condition known as prediabetes.

To address the challenge, IBM Watson Health and the ADA are collaborating to apply Watson cognitive computing to the organization's massive library of information and data. Through this effort, IBM and ADA hope to empower entrepreneurs, developers, healthcare providers, and patients to gain knowledge that can improve outcomes and even prevent the condition's onset.

The IBM-ADA collaboration was announced at the American Diabetes Association 76th Scientific Sessions event in New Orleans this week. It was one of several announcements from IBM aimed at improving research and care for diabetes patients.

Dr. Kyu Rhee, chief health officer for Watson Health, told InformationWeek that the cognitive power of Watson will be applied in the ADA collaboration in two key areas.

[Learn more about what IBM Watson Health is doing to cure the world's ills. Read IBM Watson, Heart Association Take on Cardio Health.]

First, IBM's AI platform will ingest all the medical journals, medical text books, Pub Med, and other diabetes literature and resources available, including all the content from the ADA's Diabetes Information Center.

In a blog post published this week, Rhee wrote, "By joining forces, IBM and the [ADA] will work together to apply the cognitive computing power of Watson to the [ADA]'s unmatched repository of 66 years' of clinical and research data."

Second, Rhee said, Watson will ingest the ADA's diabetes data sets. "That information is very relevant, too, to see what are the factors that predict the development of diabetes," he told InformationWeek in an interview.

Watson will be trained to understand the diabetes data to identify potential risk factors and create evidence-based insights that can be applied to health decisions, Rhee wrote in his blog post.

Diabetes App Development

IBM also shared updates about its work with particular partners in the fight against diabetes. One of those partners is Medtronic. Together with IBM the company is in the final stages of creating a cognitive app called SugarWise. The first generation of the app creates a retrospective analysis of a patient's insulin, continuous glucose monitors (which are wearables that track glucose levels 24 hours a day), and nutritional data to uncover patterns and trends, according to a prepared statement. The goal is to help people understand how their behavior affects glucose levels in real-time.

IBM and Medtronic have also created a new hypoglycemia prediction engine using data from 10,000 anonymized patients.

Meanwhile, developers at digital health startup HelpAround are building Watson APIs into the company's diabetes support community in order to optimize their ability to analyze patient questions and requests in real-time. HelpAround offers a mobile platform that matches diabetic patients with the resources they need.

Diabetic Complications Research

IBM also is collaborating with the Health Maintenance Organization Maccabi Healthcare services to build a predictive machine learning model to help identify early risks for diabetic retinopathy, the top cause of blindness for those with diabetes.

"Technology promises to transform the way doctors and patients battle disease, especially complex chronic illnesses like diabetes," said Professor Varda Shalev, director of the Institute for Health Research and Innovation at Maccabi Healthcare Services, in a prepared statement. "The goal of our work with IBM Research is to discover new ways to significantly improve patient treatment and our ability to help prevent retinal degeneration for people with diabetes."