IBM’s Watson artificial-intelligence computer system made headlines a couple years ago by besting two champion contestants on the TV quiz show, Jeopardy!, winning a $1million prize. Watson’s claim to fame was not just its ability to process vast amounts of data. More importantly, it had a humanlike understanding of language and could quickly make complex decisions to come up with the best answer.
Researchers at the recent Cleveland Clinic Foundation (CCF) Innovation Summit related how they are collaborating with IBM scientists on two Watson projects that aim to help doctors make quicker and more-accurate decisions.
If all goes well, the Watson EMR Assistant will tap the unrealized potential of electronic medical records. Currently, records for even a single patient can contain vast amounts of data, including plain-text notes, family history, lab results, past procedures, medication history, and so on. Data are usually organized in many different ways by different people and organizations. So the odds that a doctor sees all the info needed to make an informed decision are problematic at best. A cognitive computer, on the other hand, can quickly wade through the complex and disparate data and highlight pertinent details for the physician — ultimately improving the efficiency and quality of care.
The second project, WatsonPaths, is a teaching aid. When presented with a medical case, it explores various possibilities based on information from reference materials and medical journals, forms hypotheses, and recommends treatments. And it incorporates feedback from medical experts on which evidence is more important and leads to the strongest conclusions. Through this collaboration loop, WatsonPaths will get “smarter” over time.
With an emphasis on critical thinking and problem solving, WatsonPaths will help medical students understand what information sources Watson consulted, what logic it applied, and what inferences it made in arriving at recommendations.
The current aim with these technologies is to assist humans, not replace them. But cognitive machines have the potential to help diagnose diseases and assess the best treatments. “Of course, it is also easy to visualize how this type of technology could eventually be a tool for physicians to use in real-time clinical scenarios — a powerful guiding reference to consult when diagnosing and identifying the best treatment options,” says Eric Jelovsek, Director of the Cleveland Clinic Multidisciplinary Simulation Center.
So as Watson learns and its capabilities improve, the obvious question is: Will this technology eventually supplant human physicians? Not anytime soon, according to Innovation Summit presenters. There’s a lot more to practicing medicine than just rote I/O, especially interacting with patients, noting subtle signs that don’t show up in any test, and even tailoring treatment based on the person’s demeanor. No computer comes close to that today. “Watson may someday be the world’s smartest doctor, but it will never be the wisest,” noted one Clinic physician.
While that’s reassuring, a more-troubling note during the Summit was IBM’s attitude toward security. The onslaught of data-mining and now-commonplace data breaches have raised red flags for anyone who’s online, so an obvious area of concern is medical-history confidentiality. Amazingly, an IBM spokesman gave a curt dismissal to this possibility, indicating it is well under control and simply not an issue worth discussing. Maybe this is well-founded confidence, maybe it’s hubris. But with just one major data compromise, Dr. Watson could be DOA.