Watson is going to work in health care (and we'll all benefit)

Dr. Martin Kohn is the Chief Medical Scientist for IBM Research.

How about this for information overload?

Every five years, the amount of new medical information that is created doubles. Yet, 81 percent of doctors say they can spend less than three to five hours a week keeping up with it.

As a profession, we’re creating a lot of life-altering knowledge. But it’s impossible for us to memorize every technical journal and textbook available. Is it any surprise that only about 20 percent of the knowledge clinicians use today is evidence-based?

And the problem isn’t just that we need help sorting through all this information, whether it’s electronic medical records, clinical studies published in medical journals, or test results.

It’s that as this data piles up and major advances are made in research about specific illnesses, treatments, and genetics, we also need help finding the relevant information so we can diagnose illnesses faster and more accurately.  We can learn from what works with other patients in order to be able to tailor treatments better to each individual’s personal situation.

Since I am an emergency physician, I'm very pleased that the first commercial applications for IBM’s Watson are in the medical field. The powerful combination of capabilities like big data analytics and cognitive computing found in Watson stand to revolutionalize healthcare through individualized, evidence-based medicine.

Watson made a splash two years ago, when it beat the grand masters of the Jeopardy!quiz show by analyzing the meaning and context of human language and quickly processing vast amounts of information to come up with the most probable answers to questions.

But Watson isn’t playing games anymore. This time, instead of competing with humans, Watson is teaming up with us. In fact, Watson has been learning about medicine, with teams of nurses and doctors teaching it the intricacies of its new jobs.

For more than a year, IBM has partnered separately with WellPoint, a health insurance provider, and Memorial Sloan-Kettering to train Watson in the areas of oncology and utilization management. During this time, clinicians and technology experts spent thousands of hours "teaching" Watson how to process, analyze and interpret the meaning of complex clinical information, using natural language processing. After working together during the past year, IBM, Memorial Sloan-Kettering Cancer Center, and WellPoint rolled out the first Watson-based products aimed at improving the quality and speed of patient care.

One such product, designed with help from my company, helps physicians assess and manage treatments for cancer, initially lung and breast cancers. The cloud-based app, which doctors tap into using PCs, tablets or smart phones, pulls together different streams of data, including patient records, clinician notes, and test results, identifies missing pieces of data and possible tests, and suggests different treatment options.

Oncology experts have trained Watson to compare a patient’s medical information against treatment guidelines and published research, sharing their knowledge and expertise to help Watson learn from their own experiences with cancer care.

But what I think makes the technology so powerful is that it can take information about a specific patient and match it to a vast trove of information about the history of treatment of similar patients. The goal is to make it easier for doctors to have access to the collective wisdom about oncology so that it can be used to make evidence-supported decisions.

So far, more than 600,000 pieces of medical evidence, two million pages of text from 42 medical journals, and clinical trials in cancer research have been fed into Watson. Watson can sift through 1.5 million patient records to provide doctors with treatment options in seconds.

WellPoint, meantime, worked with us to develop an application that helps health plans make decisions about treatment requests more quickly and consistently in an effort to streamline the prior authorization process.

WellPoint trained Watson with 25,000 test case scenarios and 1,500 real life cases, teaching it to interpret the meaning of queries and analyze them in the context of medical data and human and natural language, including doctors notes, patient records, and clinical feedback. Nurses spent almost 15,000 hours training with Watson, which continues to learn on the job.

The data deluge and how we can used technologies like analytics to derive insight from information were among the topics that I heard discussed this week at the annual HIMSS conference, where I was one of thousands of clinicians and health IT professionals who descended upon New Orleans to collaborate and explore ways in which technology, like Watson, is changing the face of healthcare today.

This is where I think Watson will be so transformative for us in the medical profession: by helping us share the hard-won knowledge we’re working to uncover so that we can save lives, make decisions more quickly, and alleviate people’s pain and worries. Because Watson isn’t the decision maker. That’s the realm of the medical professionals. What Watson does is open up that treasure trove of information already out there to help us make better, smarter, more compassionate decisions.

For more information about Watson and health care, check out this video.

This post was originally published on Smartplanet.com