IBM Watson to Aid Sloan-Kettering With Cancer Research

By: Brian T. Horowitz

2012-03-26

Memorial Sloan-Kettering Cancer Center will use IBM Watson's supercomputer to cull its rich database of clinical knowledge to help doctors treat cancer.

IBM Watson, the physician's assistant, has another gig in cancer care. Big Blue and Memorial Sloan-Kettering Cancer Center, the world's oldest and largest cancer center, have announced plans to use the Watson supercomputer to develop a decision-support application for cancer treatment.

The arrangement will combine Watson's computational and natural-language processing abilities with the clinical knowledge of Sloan-Kettering, which has a database of more than 1 million patients.

Oncologists at Sloan-Kettering, based in New York City, will use Watson to gather patient information, treatment guidelines and published research in order to develop personalized treatment options for cancer patients. As Sloan-Kettering provides large volumes of clinical data and Watson performs its complex analysis, doctors will be able to generate and evaluate hypotheses on evidence and treatment.

A veteran contestant on "Jeopardy," Watson can search millions of pages in seconds, IBM reported. Watson uses natural-language processing (NLP) capabilities from Nuance Communications to interpret queries from doctors. NLP takes data from unstructured documents, such as doctors' notes, admittance records, research findings and journal articles, said Stephen Gold, global strategist for IBM Watson Solutions.

Watson presents data to physicians and researchers in the form of an evidence panel as it dissects and answers questions as it did on "Jeopardy," Gold wrote in an email to eWEEK.

The new tool's interface will allow doctors to drill down to the details behind cancer case histories and view the evidence that led to outcomes in those cases, Gold explained.

In addition to Watson, the decision-support system will incorporate molecular and genomic data, as well as a database of Sloan-Kettering's de-identified cancer patient data and patients' electronic health records (EHRs).

"Consistent with our mission, the vision is to help better identify and personalize cancer therapies for each individual patient, no matter where that patient may be receiving care," Craig B. Thompson, Sloan-Kettering's president and CEO, said in a statement.

Keeping up with rapidly changing data is a challenge for oncologists and general physicians, especially when they don't specialize in particular types of cancer, IBM reported.

"This comprehensive, evidence-based approach will profoundly enhance cancer care by accelerating the dissemination of practice-changing research at an unprecedented pace,” said Dr. Mark Kris, chief of the Thoracic Oncology Service at Sloan-Kettering and one of the clinicians leading the development of Watson at the cancer center.

"Cancer care is profoundly complex, with continuous clinical and scientific advancements to consider," Dr. Martin Kohn, IBM's chief medical scientist, said in a statement. "This field of clinical information, given its importance on both a human and economic level, is exactly the type of grand challenge IBM Watson can help address."

On Sept. 12, 2011, IBM announced it would work with health insurer WellPoint to develop health care applications using Watson. Then on Dec. 16, IBM said Cedars-Sinai's Samuel Oschin Comprehensive Cancer Institute in Los Angeles would advise WellPoint on how to create applications using Watson to help doctors develop cancer treatment plans.

In late 2012, oncologists at Sloan-Kettering will begin pilot projects using Watson to develop treatment plans for breast, lung and prostate cancers, of which Sloan-Kettering's oncologists are sub-specialists. The health organization will expand its distribution of the technology in late 2013. Other areas could include wellness care, cardiology and managing chronic conditions.

"This is still the early days for Watson, but IBM is certainly hoping to advance Watson in other areas," said IBM's Gold.

More Prof. Ravi K. Sharma's questions See All
Similar questions and discussions