Experts in health care and information technology agree on the future’s biggest opportunity: the creation of a new computational model that will link together all of the massive computers that now hold medical information. The question remains: who will build it and how?
IBM is today staking its claim to be a major player in creating that cloud and to use its Watson artificial intelligence – the one that won on the TV game show Jeopardy – to make sense of the flood of medical data that will result. The new effort uses new, innovative systems to keep data secure, IBM executives say, even while allowing software to use them remotely.
“We are convinced that by the size and scale of what we’re doing we can transform this industry,” says John Kelley, Senior Vice President, IBM Research. “I’m convinced that now is the time.”
Big Blue is certainly putting some muscle into medicine. Some 2,000 employees will be involved in a new Watson-in-medicine business unit. The Armonk, N.Y.-based computing giant is making two acquisitions, too, buying Cleveland’s Explorys, an analytics company that has access to 50 million medical records from U.S. patients and Dallas’ Phytel, a healthcare services head of IBM’s Life Science company that provides feedback to doctors and patients for follow-up care. Deal prices were not disclosed.
It is also announcing some big partnerships:
Apple will work to integrate Watson-based apps into its HealthKit and ResearchKit tool systems for developers, which allow the collection of personal health data and the use of such data in clinical trials.
Johnson & Johnson, which is one of the largest makers of knee and hip implants, will use Watson to create a personal concierge service to prepare patients for knee surgery and to help them deal with its after effects.
Medtronic, the maker of implantable heart devices and diabetes products, will use Watson to create an “internet of things” around its medical gadgets, collecting data both for patients’ personal use and, once it’s anonymised, for understanding how well the implants are working. Initially, the focus is on diabetes.
IBM’s pitch is that it will be able to create a new middle layer in the health care system – linking the old electronic records systems, some of which have components dating back to the 1970s, with a new, cloud-based architecture, because of its deep breadth of experience.
And there is no doubt that there is a need for data science that can parse the explosion of information that will soon be created by every patient. Already, there is too much information for the human brain. “If you’re an oncologist there are 170,000 clinical trials going on in the world every year,” says Steve Gold, VP, IBM Watson.
The question is how ready Watson is to take on the challenge. IBM isn’t the only one that sees opportunity here. The billionaire Patrick Soon-Shiong is aiming to create a system to do many of the same things with his NantHealth startup. Flatiron Health, a hot startup in New York, is creating analytics for cancer.
The existing health IT giants, Cerner and Epic, both certainly have their eyes on trying to capture some of this new, interconnected market, lest it make them obsolete.
So far, Watson has been a black box when it comes to healthcare. IBM has announced collaborations with Anthem, the health insurer, and medical centres including M.D. Anderson, Memorial Sloan-Kettering Cancer Centre, and The Cleveland Clinic.
There are lots of positive anecdotal reports, but so far the major published paper from Watson is a computer science paper published by the Baylor College of Medicine that identified proteins that could be useful drug targets.
“I think that ultimately somebody’s going to figure out how to integrate all these sources of data, analyse them, sort the signal to noise, and when someone can do that, it will improve the health care system,” says Robert Wachter, the author of The Digital Doctor: Hope, Hype, and Harm at the Dawn of Medicine’s Computer Age and associate chair of medicine UCSF.
“Does this do that tomorrow? No. But do we need to create the infrastructure to do that? Yes. And are they probably the best-positioned company with the best track record to do this? I think so.”