on Tuesday, June 5, 2012
Memorial Sloan Kettering and IBM have agreed to collaborate on the development of a powerful tool built upon IBM Watson to provide medical professionals with improved access to comprehensive cancer data and practices.
This is the same IBM Watson that gained fame by beating human contestants on the TV game show Jeopardy in 2011. The computer is able to interpret queries in natural language and uses statistical analysis, advanced analytics, and a powerful array of processors to search millions of pages in seconds and deliver evidence-based, statistically ranked responses.
The joint project will combine the computational power of IBM Watson and its natural language processing ability with the clinical knowledge of Memorial Sloan Kettering’s physicians, the Center’s existing molecular and genomic data, and its vast repository of cancer case histories — along with continuously updated standards of care and other published data — to build the world’s most complete, cancer-specific medical resource tool. Memorial Sloan Kettering oncologists will “teach” IBM Watson to evaluate and synthesize the enormous stores of clinical information and the collective wisdom and experience of the Center’s cancer experts, all of which will be put into the system.
Insights gleaned from Memorial Sloan Kettering clinicians’ years of experience will be incorporated into individualized treatment recommendations; in addition, IBM Watson will provide physicians with a record of the data and evidence that were used to reach those recommendations. “The combination of transformational technologies found in IBM Watson with our cancer analytics and decision-making process has the potential to revolutionize the accessibility of information for the treatment of cancer in communities across the country and around the world,” says Memorial Sloan Kettering President Craig B. Thompson. “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. Our clinicians and researchers can also expect next-generation research opportunities to emerge as we use the powerful analytics provided by IBM Watson to address innovative clinical research issues.”
The need for such an advanced technology arises from the increasing complexity of cancer treatment and the fact that approximately 85 percent of cancer patients today are not treated at specialized medical centers. Oncologists and physicians who do not specialize in treating specific subtypes of cancer face significant challenges in keeping up with the wealth of rapidly evolving cancer information. It can take years for the latest research and guidelines to be circulated to the thousands of community-based oncologists who treat most patients. The IBM Watson tool is expected to speed this process.
“This comprehensive, evidence-based approach will profoundly enhance cancer care,” says Mark G. Kris, Chief of the Thoracic Oncology Service and one of the Memorial Sloan Kettering clinicians who led the development effort, along with Larry Norton, Deputy Physician-in-Chief for Breast Cancer Programs; Howard I. Scher, Chief of the Genitourinary Service; pulmonary and critical care physician and epidemiologist Peter B. Bach; and Memorial Sloan Kettering ’s Chief of Strategic Planning and Innovation, Wendy Perchick. “It will accelerate the dissemination of practice-changing research at an unprecedented rate,” Dr. Kris adds. The initial focus will be on lung, prostate, and breast cancers with a goal of piloting the tool to a select group of oncologists in late 2012. Wider distribution is planned for late 2013.
“Memorial Sloan Kettering is a data-rich organization,” explains Patricia C. Skarulis, Vice President of Information Systems and Chief Information Officer. “One of the biggest appeals of this project for both Memorial Sloan Kettering and IBM is to take advantage of the extraordinary data warehouse that Memorial Sloan Kettering has built over the past 25 years that integrates all relevant [de-identified] clinical information on our patients.”