Watson could transform medical care
Seldom do I watch the Jeopardy! game show anymore; however, the promos for the IBM super-computer’s match against Jeopardy!’s top-two money winners, Ken Jennings and Brad Rutter, hooked me. As it turned out, their brains were no match for the room-sized Watson with scores that fell $50,000 short of his…uh, its…final tally.
The creation of this machine that “understands” human language fascinated me more than its warp speed and relative accuracy Maybe it’s my leftover fascination with Star Wars’ C3PO, but whatever the source of my curiosity, I’m glad I tuned in. I witnessed the real-life promise of what is to come.
Watson does more than just regurgitate data. It makes connections among words and phrases to determine a “logical” answer from that data, analyzing the type of wordplay so familiar to Jeopardy! aficionados. Watson also provides you the statistical probability of a correct answer.
Admittedly, some of Watson’s responses made me feel as if I was listening to an English translation of a Japanese instruction book—answers that were laughably off base but close enough to understand the intended connection. Watson’s much-publicized flub at the end of the first day—when the computer’s response was to have been a U.S. city and he named Toronto—reminded us that Watson is, after all, just a machine, although one capable of warning us of a high probability of an incorrect answer.
Shortly after the broadcast, IBM officials announced that Watson will now go to medical school, two of them: Columbia University and the University of Maryland, where Watson’s power will be tapped as a diagnostician, helping doctors identify diseases and recommend treatments. Doctors are more likely than not to input patient data into their records digitally these days, and they receive digital lab test reports. Watson can store and collate this patient data, as can many computers, but now that information can also be interpreted in a matter of seconds, analyzing patient information along with published research from medical journals worldwide—hundreds and hundreds of pages—while calculating the probability of each diagnosis at the same time.
I believe my dad, an old-time country doc in Homestead, was the “Watson” of his day. Often he was sought out by fellow physicians for his “uncanny” diagnostic ability. He read a book daily, sometimes more, and he used all his tactile senses to aid his diagnoses. “Smell that?” he’d say to me, as I tagged along on house calls to patients’ homes, toting his black leather doctor’s bag with both hands. Hmm, I smelled blueberry muffins or apple pie at times, but nothing of the kinds of human odors that spoke to him.
On occasion I’d see him lick a small patch of a patient’s arm, seeking those elusive clues to what was wrong, or to confirm what he had suspected.
Other times, he would sit wordlessly beside a patient’s bed, holding one hand in both of his after he had listened to a heart and lungs struggle to beat, to breathe. It seemed to me he was transfusing those patients with his own life energy; and they always responded, no matter how ill, sometimes opening eyes that had closed weeks earlier.
Critics of Watson’s introduction into medicine charge that a computer will likely put more distance between a patient and her doctor. If, perhaps, the old country doc was still alive and the norm, rather than the exception, I might be convinced. Today’s reality, however, quite often minces patient symptoms into ever-more-specialized “specialities” requiring ever-more-qualified specialists, strangers caring for nearly nameless patients, their questions or answers misunderstood through the sieve of often unfamiliar accents.
How appealing to make an appointment with Dr. Watson, whose questions would be offered in his monotone-but-mellifluous voice, who could diagnose your ailment within seconds, perhaps churning out a legible prescription and treatment course at the front desk with only 15 minutes taken from your day. A nearly perfect arrangement, it seems to me … as long as they can make Watson look a little like George Clooney.