IBM's Watson, a question-answering computer that uses semantic understanding to respond to questions posed in natural language, is beginning to see real-world use as cancer diagnosis intelligence in 21 different hospitals across China.
What Is Watson?
IBM developed the Watson computer, named after IBM's first CEO Thomas J. Watson, to explore the possibility of data-analysis-based computed responses to natural language input. The computer can learn through analyzing and parsing data input related to the topic it will be researching, computing, and for which it will be producing answers.
As a test of Watson's ability to perform its intended task, creators pit the computer against human opponents on the trivia game show Jeopardy! where Watson had to perform deep content analysis, information retrieval, and natural language processing by utilizing machine learning and artificial intelligence.
Watson claimed the first place prize of $1 million US even faced with an unpredictable and broad domain of knowledge. Watson's success opened the door to more philanthropic work, including piecing together accurate diagnoses for potential cancer patients.
What Will Watson Do in China?
The number one cause of death in China is cancer. It's responsible for nearly a quarter of all deaths in the country. Of the developed countries that may need the help of more accurate cancer diagnoses, China is at the top of the list. IBM has partnered with Hangzhou CognitiveCare to learn Watson’s potential as a cancer diagnostic tool in a high-risk cancer environment.
The goal of employing Watson in 21 rural and urban hospitals in China is to turn over 200 textbooks, 300 medical journals, and almost 15 million pages of medical text into personalized cancer treatments that also consider the medical history of patients.
Watson may be able to help doctors more accurately diagnose patients and provide better treatment solutions.
Source: va.gov
This iteration of Watson is called Watson for Oncology. IBM computer scientists developed the specialized version of Watson with help from oncologists at Memorial Sloan-Kettering Cancer Center in New York. They designed Watson for Oncology to read data from both structured and unstructured sources so that hospitals without digital libraries can still provide the computer with vital information.
In the future, IBM hopes to localize Watson so that it operates in Chinese as well as English. If the computer meets with success, IBM and Hangzhou CognitiveCare will begin making efforts to bring the computer to as many in-country oncologists as possible.
As of now, Watson has helped cancer diagnosis reach about 90% accuracy. While 100% accuracy is ideal, 90% is a large leap from the approximate 50% accuracy that physicians and oncologists had previously achieved. With a small amount of additional research and some positive results in the field testing in China, Watson’s cancer diagnosis abilities may become the next essential tool in the ongoing fight against all types of cancer that cause so many deaths in developed and undeveloped countries.