With Watson, IBM Seeks to Sell Medical Knowledge
On the TV show House, Dr. Gregory House spends most of each hourlong episode wrestling with how to diagnose a patient who presents a bewildering set of symptoms.
IBM research engineer Steve Daniels jokes that he and his colleagues could turn House into a “five-second show.” The doctors would simply ask, “Hey, Watson, what does this guy have?”
Watson is the supercomputing engine that beat the top two human competitors on the quiz show Jeopardy! this year, and Daniels is on the IBM team developing the software’s first commercial application as what could be a stunningly useful diagnostic assistant for doctors. If it works as envisioned, Watson could help doctors identify what is afflicting any patient and suggest a course of treatment.
With Watson, the company is also experimenting with a new business model: knowledge-as-a-service. Within two years, IBM says, it will begin charging health-care practitioners to log in to computers running the Watson software so they can get help cracking tough medical questions or finding the most cost-effective treatment.
A “Dr. Watson” would not be the first automated diagnostic tool; other online systems are already available to walk physicians through lists of suggestions for what might be ailing a patient. However, IBM researchers believe Watson will outshine the others because it can automatically digest not only textbook medical knowledge (the software can read one million books in three seconds) but also patient blogs, insurance claims, and millions of observations that doctors are increasingly being asked to encode into electronic patient records.
IBM says Watson can’t replace doctors, yet behind Watson and other automated diagnostic tools is a recognition that no human can keep up with expanding medical knowledge. New treatments and new subcategories of disease are constantly being discovered, and even long-documented illnesses might be so rare that most doctors would never have encountered a case before.
Studies have found that doctors misdiagnose patients 10 to 15 percent of the time, sometimes with fatal consequences. Herbert Chase, a professor of clinical medicine at Columbia University who is working with IBM to turn Watson into a diagnostic advisor, is still haunted by a case from 35 years ago. He encountered a patient with rickets, which causes bones to soften. He and his colleagues were mystified about the cause, because the patient didn’t have vitamin D deficiency, the usual culprit. It took four months before Chase came across a reference in the medical literature to a rarer cause of rickets, one related to a kidney malfunction. The patient survived, but she suffered more than she should have. “I’ve spent my whole life in medicine grappling with how hard it is to find information that is known,” he says.
A family medical drama motivated Jason Maude to launch Isabel Healthcare, which runs an online diagnostic-support service used by about 30 U.S. hospitals and medical practices. In 1999, Maude’s three-year-old daughter, Isabel, was being treated for chicken pox but ended up suffering a multisystem organ failure. Doctors were stumped about how to treat it; finally they realized she had necrotizing fasciitis, a severe bacterial infection. She barely survived, and Maude quit his job in finance to launch a service that could help doctors spot mysterious illnesses sooner, improving a patient’s odds and decreasing the need for unnecessary tests. Today Isabel is available for individual doctors for $375 per year; a community hospital would pay about $50,000 annually, and a large medical system as much as $400,000.
The system does some of what IBM envisions Watson will do. A doctor can log in to Isabel over the Internet and enter a patient’s age, gender, and symptoms, using either medical terminology (“hand edema,” for example) or everyday language (“swelling of the hand”). Isabel will display a list of likely diagnoses, with links to supporting evidence in medical textbooks and journal articles. As more symptoms are added, Isabel recalculates the likelihood of each diagnosis. For example, entering that a 30- to 39-year-old man has swollen hands and fatigue leads Isabel to suggest hypothyroidism as the likeliest cause, followed by less probable ones from kidney ailments to carpal tunnel syndrome.
Both Isabel and another diagnostic tool called SimulConsult, which specializes in helping doctors diagnose metabolic and genetic disorders, use information painstakingly organized by teams of medical specialists. But Watson won’t be loaded that way, which could be a strength or a weakness. Before IBM put the computing engine on Jeopardy!, its creators didn’t feed Watson all the possible clues and answers that might come up on the show. Instead, they encoded rules about English grammar into the software, then fed the program with unstructured information—massive amounts of text—and let Watson make its own connections between words and phrases that tended to appear together. From these statistical correlations, Watson inferred facts. No one specifically told the program that Theodore Roosevelt negotiated the Treaty of Portsmouth, but Watson could guess after seeing countless pages about the Russo-Japanese War.
Similarly, IBM researchers are exposing Watson to medical textbooks and journal articles. It’s also seeing anonymized medical records held by WellPoint, a large U.S. health insurance company that plans to eventually have Watson analyze patient records and then give its suggestions to doctors. Someday, Watson could also be pointed to blogs or other websites where patients talk about their responses to medical treatments.
Watson still needs extensive testing before IBM can begin selling its services in the next year or two. That’s why IBM is working with doctors, including Chase and his students at Columbia, who pose questions to the program to see how well the computer answers. IBM can then fix errors they spot—like the time the computer included penicillin’s discoverer, Alexander Fleming, in a list of antibiotics.
Maude says he thinks such glitches will always plague Watson. “You can do something automatically, but the accuracy and the relevance will go down,” he says. “We’ve found that’s often where the wild things come in.” Michael Segal, the doctor who founded SimulConsult, similarly argues that there are instances when a computer can be tripped up by nuances in medical terminology. For instance, he says, the term “congenital” generally means present at birth. But in the context of neuromuscular disorders, “congenital” can be used to describe something that has its onset in the first two years. “It’s just a nightmare to try to do this in an automated way,” he says.
Chase says that even with its occasional hiccups, Watson outperforms any other diagnostic software he has tried, not only because of its broad knowledge base but because doctors can ask it open-ended questions in “natural language,” such as “What is the recommended antibiotic for a pregnant woman with Lyme disease and a penicillin allergy?” IBM is also working with Nuance Communications, a maker of voice recognition software, so that doctors will be able to speak their queries to Watson rather than having to enter them at a keyboard.
But getting the technology right is just part of the project. Resistance from doctors could be a bigger hurdle to getting Watson into widespread use; Chase says he has encountered “insecurity” from doctors when he has talked to them about the program. In trying to sell Isabel, Maude stresses to physicians that while software can help them make a diagnosis, it can’t do it for them. But even that makes some of them bristle. “They do think it takes away some of the magic of being a doctor,” he says. “The ability to diagnose—for them it’s their core skill. It’s kind of what their peers most revere about them.”
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
ChatGPT is going to change education, not destroy it
The narrative around cheating students doesn’t tell the whole story. Meet the teachers who think generative AI could actually make learning better.
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.