In partnership with the American Cancer Society (ACS), IBM developed a tool based on Watson that uses machine learning to provide patients with information and personal advice about the disease. Initially, the counselor will advise on the type of cancer the patient has, on the stage of the disease and treatments that were administered.
From these and other data, it tries to give advice to the patient and answer questions. With voice recognition capabilities and natural language understanding to Watson, users can ask questions and have spoken responses.
For example, a person with breast cancer may ask the adviser to explain the reasons for the pain she feels at some point. In reply, the tool will be based on what Watson learned, including the experiences of other patients with the same symptoms, and offer personalized care solutions based on the patient’s situation. The more it knows about the patient, the more the advisor can refine the user request and recommendations correspond more to the preferences of the patient.
To create the tool, IBM and the American Cancer Society will exploit their massive data warehouses and use them to form Watson. Among these data, the supercomputer will also have access to 14,000 pages on the site cancer.org where it can find detailed information on more than 70 types of cancer, more aggregate data from the National Cancer Information Center of the ACS where it will find information on the care, support groups, activities, wellness and education on cancer.
The initiative IBM is not unique. Other players in the technology industry are also seeking to facilitate the sharing and analysis of large amounts of data to improve patient care and advance therapeutics. This is the case for example with Intel’s project Cancer Collaborative Cloud, which should enable hospitals and universities to more easily share genomic data, imaging and clinical patient data for research purposes.