IBM Watson Imaging Collaborative Adds 16 Health Systems, Tech Firms

IBM Watson Health is casting a wider net to feed its artificial intelligence system, forming a 16-member international collaborative of health systems, academic medical centers and imaging technology companies to speed development of products that diagnose patients correctly the first time.

From hospitals in rural America and academic medical centers in the inner city to imaging companies in Europe and Japan, IBM said the collaborative broadens its effort to identify healthcare’s big problems in “real-life” settings in daily practice. The Watson Health business already has data on an estimated 300 million patients and the collaborative will help capture the experience of doctors and imaging centers working with different diseases and imaging equipment used to treat patients.

Steven Tolle, chief strategy officer for Watson Health Imaging, said in an interview the collaborative will help doctors address several cancers including breast and lung as well as diabetes, eye health and brain and heart diseases. And Tolle assured IBM Watson’s observers that the collaborative will be broader than its rivals.

“They partner with one partner and go after one disease,” Tolle said of IBM Watson rivals. “That’s not good enough. We are going to go after 16 partners and go after the entire body, starting with the diseases that cost the most and kill the most.”

IBM Watson Health wants to be a major player in improving health outcomes as the U.S. healthcare system moves away from fee-for-service medicine to value-based care. To help better manage populations of patients through the Watson health platform, the company said it needs to draw data and insights from a “variety of patient care environments ranging from stand-alone ambulatory settings to integrated delivery networks,” the company said in a statement announcing the collaborative.

The collaborative is designed to provide IBM Watson a “real-world experience and share real-world findings” because one hospital may use a different imaging machine, operational approach or medical protocol. IBM needs a diverse group of imaging experts training Watson and the collaborative will provide that.

The initial partners include: UC San Diego Health, University of Miami Health System, University of Vermont Health Network and Eastern Virginia Medical School, as well as imaging firm Agfa HealthCare, Japanese eye care device maker Topcon, the teleradiology provider vRad and Hologic , an imaging system developer. Also involved is IBM Watson Health unit Merge Healthcare, a company that helps doctors and hospitals store and analyze CAT scans and X-rays, putting medical imaging into a cloud.

“We want to get all kinds of source data to train Watson,” Tolle said. “We want to get to a point where you could turn this on for a rural community and academic medical center and get the same level of accuracy and specificity.”

IBM already has partnerships with CVS Health , Johnson & Johnson and Medtronic to better predict patient health and link to its cloud-based architecture.

“With an ability to draw insights from massive volumes of integrated structured and unstructured data sources, cognitive computing could transform how clinicians diagnose, treat and monitor patients,” Anne Le Grand, who recently joined IBM as vice president of Imaging for Watson Health, said in a statement. “Through IBMs medical imaging collaborative, Watson may create opportunities for clinicians to extract greater insights and value from imaging data and better manage cost.”