How Watson Is Changing the Course of Medical History
IBM’s brain child, Watson, has been best known as defeating humans on the game show Jeopardy in 2011 where it competed against Brad Rutter and Ken Jennings, two of Jeopardy’s most famous contestants. Watson was made specifically to play Jeopardy by its research team headed by David Ferrucci of IBM.
Since its debut on Jeopardy, Watson has been used in a number of different fields such as healthcare and financial services to improve the decision making process. The sheer volume of information that is available is growing at such a fast rate that it is humanly impossible to keep up. It is predicted that the rate of data growth will be upwards of 800 percent in the future 5 years. Many people have suggested that computers should be able to help with this problem but in reality, current systems have prevented real growth in this area. Natural language is very intricate. It is often implicit: the exact meaning is not completely and exactly stated. In human language, meaning is highly connected to what has been said before, the topic itself, and how it is being discussed: factually, figuratively or fictionally – or a combination.
How do individuals, professionals and organizations stay up-to-date? This has been a major problem in professional fields like healthcare, where anyone that is connected to healthcare cannot stay informed on the latest best practices due to the huge scope that healthcare consists of. Due to this gap in information available versus information used properly, it has been suggested that as high as 20% of medical diagnosis is incorrect. This affects doctors and their ability to diagnosis a problem as well as healthcare providers such as insurance agencies that need the best and most accurate information to process claims and still operate as a business.
With all the most current information, this will be able to change the way that the healthcare world operates today.
The information that medical professionals need to make accurate decisions is available to them now. Medical journals publish new treatments and uncover new discoveries every day. Using patients history records also give a clue. All of this is great but due to the large amounts of electronic medical record data, it is impossible to keep up.
Enter Watson into healthcare
In February 2013 IBM Corp., the health insurance provider WellPoint Inc. and Memorial Sloan-Kettering Cancer Center announced two Watson-based applications. The first is to help assess treatment options for lung cancer and second is to help manage health insurance decisions and claims. In both applications, doctors or insurance company workers will access Watson through a tablet or computer. Watson will quickly compare a patient’s medical records to what it has learned and make several recommendations in a decreasing order of confidence. In the cancer program, the computer will be considering what treatment is most likely to succeed. In the insurance program, it will consider what treatment should be authorized for payment.
Watson provides natural language processing, natural language understanding, machine learning, hypothesis generation, probabilities of various outcomes, and evaluation to give direct, confidence-based answers. Watson can learn from success and failure to improve future responses. Watson can take that information and turn it into the insight needed to avoid improper diagnosis or unnecessary treatment. This breaks down the barriers between people and machines.
Manoj Saxena, an IBM general manager, said the supercomputer has ingested 1,500 lung cancer cases from Sloan-Kettering records, plus two million pages of text from journals, textbooks and treatment guidelines.
Given the growing complexity of medical decision making, Watson has been targeted by WellPoint to help streamline their process regarding insurance claims.
WellPoint will be using the system internally for its nurses and clinicians who handle utilization management. The company will also make the intelligence available as a Web portal to other providers as its Interactive Care Reviewer.
Watson will support medical professionals as they make decisions. A physician can use Watson to help in diagnosing and treating patients. First the physician might pose a query to the system, describing symptoms and other related factors. Watson then goes through an inquiry process to pull pertinent information. Watson then searches patient data to find relevant facts about family history, current medications and other conditions. It than puts together this information with its findings from tests. The next step is Watson examines all available data sources to form hypotheses and test them. Watson can incorporate treatment guidelines, electronic medical record data, doctor’s and nurse’s observations, research, clinical studies, journal articles, and patient information into the data available for analysis.
Watson will then provide a list of potential diagnosis along with a score that indicates the level of confidence for each hypothesis.
The process is very similar for insurance use. Watson will come to a hypothesis for the treatment chosen to be the most relevant in order to process for payment.
The benefits Watson can provide for a healthcare provider is invaluable and will change the future of medical history in such a positive way for everyone involved.