Elementary, my dear Watson!

25 July 2013 / California Workers Compensation, Chart Reviews, Cloud Based Computing, Impairment Physical Exam, Impairment Rating Specialists, Medical History, Medical Technology / Comments Off on Elementary, my dear Watson!

There’s a new cloud-based computer service that takes in patient information and then spits out useful results in record time. And no, I’m actually not talking about RateFast. Did you catch the 2011 Jeopardy special where Watson (IBM’s language savvy super-computer) competed against two former Jeopardy champions? The AI system out-answered—or rather, out-asked—both of its human opponents, and won $1 million in prize cash.

But Watson’s verbal talent is now taking on questions that are tougher than those Alex Trebek would ask: the computer system is diagnosing and generating treatment plans for cancer patients. An agreement between IBM, healthcare company Wellpoint, and Memorial Sloan-Kettering Cancer Center is going to make Watson’s thinking power available to hospitals and clinics around the nation.

Here’s how it works: you’re a doctor, and you need to write a treatment plan for your patient, but you’re stretched for time, knowledge, resources—or maybe you just want a second opinion. So you log into Watson using an app on your tablet or computer, and you enter your patient’s medical situation. Within minutes—usually within seconds—Watson gives you a series of treatment plans based on the latest medical research, and each plan is ranked by its expected effectiveness and cost.

Over the past few years, Watson has become an expert oncologist. According to an article at Wired Magazine, Watson needs only a few seconds to search through “600,000 pieces of medical evidence, more than two million pages from medical journals” and 1.5 million patient records. Wellpoint clams that Watson can correctly diagnose lung cancer 90% of the time, as opposed to the relatively paltry 50% correct-diagnosis rate of a human doctor.

For now, Watson’s expertise is restricted to lung, prostate, and breast cancer, but the computer never stops learning. It’s difficult for doctors to stay abreast of the current medical research in their field, but Watson can “read” the results of new studies as they’re published, and instantly apply the information.

The Memorial Sloan-Kettering Cancer Center in New York says that it would take 160 hours of reading a week for a doctor to stay current on all new medical knowledge as its published—and that doesn’t even account for applying that knowledge (or taking care of tedious clerical work like filling out PR-4 reports). Obviously, no human medical worker can do what it takes to know everything about their work. But Watson and computer systems like it are unencumbered by the human weakness of, you know, having a life.

The goal is to have computers crunch numbers and negotiate the rules of today’s labyrinthine healthcare system, while medical professionals can leave work on time. Watson could be a major wave in medical technology’s recent move toward cloud-based apps that aim to streamline productivity around hospitals and clinics.

Back to top