After conquering puny humans Ken Jennings and Brad Rutter and winning a total of $77,147 over three days and two full games on Jeopardy!, IBM's know-it-all new supercomputer is going to med school. On Wednesday, IBM, along with Nuance Communications Inc. and the Columbia University and University of Maryland medical schools, announced that they are developing Watson as a diagnostic tool that can help doctors identify diseases and recommend treatments. They hope to begin lab tests as early as next year, with real world testing later in 2012.
"What makes Watson unique is that it can rip through massive amounts of information and give a small amount of possible answers with levels of confidence," says Dr. John Kelly, IBM's senior vice president of research.
Doctors have long relied on technology to help them manage patient care electronically stored patient histories, digital lab results and machines that regulate medication are all commonplace in today's hospitals. Indeed, the first attempt to create a machine that could help diagnose human illness came back in the 1970s, when Stanford University researchers developed MYCIN a computer designed to indentify different types of bacteria responsible for infections. But even the most up-to-date systems, which were developed in the 1980s, still require physicians to spend costly time typing in test data and patient information, and still only cover a limited number of diseases.
That's why doctors like Eliot Siegel, a professor and vice chair at Maryland's department of diagnostic radiology, says Watson's capabilities are necessary now. Imagine a supercomputer that can not only store and collate patient data but also interpret records in a matter of seconds, analyze additional patient information and research from medical journals and deliver possible diagnoses and treatments, with the probability of each outcome precisely calculated. "I think it's going to usher in the next generation of medicine," says Siegel. "It takes me 20 minutes to an hour or more to read through a patient's electronic medical record. Having a computer understand and present the information to me is a huge step towards allowing me to make a better diagnosis. It is really the future of medicine."
Watson's developers have always had higher goals for the room-sized, multimillion dollar supercomputer than just winning a game show. Its ability to understand natural language makes it a valuable tool in many different applications. Unlike even the most advanced Internet search engines, which can only find results for specific requests, Watson can make connections between words and determine a logical answer from imputed data. For example, if it was given the Jeopardy! clue "This is where Stefani Germanotta was born," it could infer from the data in its memory banks that where a person was born is also known as a birthplace, and that Stefani Germanotta is actually the real name of Lady Gaga. From the statements "Lady Gaga's birthplace was in Manhattan" and "The singer of 'Born This Way' was born in the Big Apple,' Watson can correctly infer the answer New York City. The supercomputer's ability to recognize the links and associations between terms in different contexts can be further applied to the medical field, especially in the case of doctors who abbreviate or misspell terms and for patients who might not know the correct scientific term for their ailing body parts.
"It's a place where we could do real good," says David Ferrucci, IBM's principal investigator of the Watson project. "It's both an important business and an area where we can help society and help people we know. There's a crisis in this country and in the world of delivering high quality health care."
That's why going on Jeopardy! made sense. Any computer can play trivia games, but Jeopardy!'s emphasis on puns, wordplay and brain-teasers allows Watson to show what it can do in a basic way that average viewers can instantly understand. "I knew the potential was there for a great computer system that could play the game [but] I didn't give it the kind of serious thought that I should have, in terms of examining the technology that was required," Jeopardy! host Alex Trebek told TIME Techland. "It wasn't until I saw the computer play that I thought, 'Holy smokes, this is serious stuff.'" Trumping his competition Jennings and Rutter, who only earned $24,000 and $21,600 respectively, it was evident that Watson could not only recall information at lightning speed, but he could interpret the English language and more than hold his own against warm-blooded competitors when it comes to analyzing wordplay.
But TV is one thing; real life is another. Some medical professionals, including Siegel's colleagues, worry that a future Doctor Watson might make us too dependent on technology. A human diagnostician immediately understands that when we say we've got stomach pains, we could really be talking about any number of organs in the abdominal area, not just the stomach specifically; computers tend to think more literally. That's why the IBM team insists that Watson can never supplant doctors completely. Katharine Frase, vice president of industry solutions at IBM Research, envisions a future where a version of Watson can be used to assist doctors in small practices where there may not be a cardiologist or urologist on call. Clinicians can use it to get answers faster rather than spending the time looking for a specialist. With a growing number of medical studies being published every day, it's hard for doctors to keep up with all the latest data. Watson can store all that information and use it to help a doctor make his or her decision. Siegel suggests thinking of Watson as one of the other doctors on the Fox medical drama House: while it's Dr. House who always comes up with the final answer, his team provides the hints and clues that help him along the diagnostic path. Frase points out that while Watson can be taught to understand that humans can exaggerate or downplay their symptoms, a computer can't judge if patients are lying as well as a human doctor can just by looking at their faces. "I don't think that any machine is ever going to take the place of the decision making process of the human or the understanding of the consequences of one decision over another," she adds. "That's one reason why people go in person to a doctor. We've got a long way to go before a computer can read human emotion."