Cancer is the second most common cause of death in the United States, and, according to the American Cancer Society, more than 1.6 million new cases are expected to be diagnosed this year. Discoveries in molecular biology and genetics in recent years have produced new insights into cancer biology, but these advances have also ratcheted up the complexity of diagnosing and treating each case.
The disease is one of the most important fields of medicine, yet it's devilishly complex and there's too much information for any single practitioner to keep up with.
A collaboration announced today between Memorial Sloan-Kettering Cancer Center in New York City and IBM could revolutionize how physicians in the United States and worldwide get access to world-class information about cancer.
Our two organizations are combining IBM Watson's natural language processing and machine learning capabilities with Memorial Sloan-Kettering's clinical knowledge and repository of cancer case histories. We aim to develop a decision support tool that can help physicians everywhere arrive at individualized cancer diagnostic and treatment recommendations for their patients based on the most complete and up-to-date information.
I credit leaders at Memorial Sloan-Kettering for envisioning a way to have a huge impact on cancer treatment worldwide. Patricia Skarulis, the organization's chief information officer, first approached us last April, shortly after she watched the Watson computer defeat two past grand-champions on the Jeopardy! TV quiz show. She said MSK had collected more than a decade's worth of digitized information about cancer-including treatments and outcomes for all of their patients-which could be mined for insights and made widely available.
She thought Watson could help. We decided to work together to try to make that happen. And, today, we believe the goal is attainable.
Since Watson's television victory last year, IBM has been on a path to improving the technology. We're making it possible for people to engage Watson in ongoing dialogues aimed at surfacing the most useful insights. After receiving an initial query, Watson will be able to ask for additional information to help it understand more precisely what the human wants to know. Also, people will be able to view the logic and evidence upon which Watson makes a recommendation.
Memorial Sloan-Kettering's oncologists will assist in developing IBM Watson to use a patient's medical information combined with a vast array of medical information-including an extensive library of medical literature, diagnosis and treatment guidelines, a database of MSK cancer cases and the institution's knowledge management system. Watson will learn from its encounters with clinicians. It will also get smarter as it amasses more information and correlates treatments with outcomes.
Our two organizations will spend most of this year loading Watson with information. This data will be used to train a version of Watson created specifically for this task. Then, starting late this year and continuing in 2013, we'll run a pilot program focused on the diagnosis and treatment of a handful of cancers, including lung, prostate and breast cancer.
Memorial Sloan-Kettering is one of the most accomplished cancer treatment centers in the world. But, when you do the math, you see that only a small percentage of cancer patients are able to receive care at MSK and other world-renown institutions.
The vast majority of patients are treated by physicians who don't have access to the more advanced knowledge that MSK oncologists possess. If MSK and IBM succeed at developing an effective decision-support tool, physicians anywhere could potentially have access to the knowledge of some of the field's top experts-and more cancer patients could get better care no matter where they live in the world.
Can you think of other fields where IBM Watson could help bring specialized expertise to the masses?
Here's a link to a post about the hookup between IBM and WellPoint, the giant health benefits company, which is complimentary to the MSK relationship.