An Oncology Consult -- With a Computer?

Mark G. Kris, MD


April 25, 2013

This feature requires the newest version of Flash. You can download it here.

This is Mark Kris from Memorial Sloan-Kettering, speaking today about developments in a whole new area for oncology. It is the use of advanced cognitive computing in the office to help us make better decisions for patients. It allows us to take advantage of advances in computing to not just improve the choice of treatment for individual patients, but also to improve and hasten the research process.

Here at Memorial Sloan-Kettering, we have partnered with IBM and WellPoint to develop an oncology advisor system that takes advantage of the IBM Watson system. IBM Watson is 3 things. It is a massive parallel computer, extremely powerful in terms of speed and memory. It is also a very sophisticated system to read natural language. Those of you who watch Jeopardy! know that the Watson computing system received the questions as a text message and understood them. It can read natural language and text.

The third thing is that these new generations of computers are in the area of "cognitive computing," in which a computer can learn from the processes and from the mistakes it makes -- it can learn and be taught. This area is growing tremendously. Many organizations, such as AETNA and eviti Inc., are developing the same technology. It has great promise, and increasing numbers of people are interested in this area. We are going to see this competitive environment making things happen even quicker.

I want to acknowledge that after a year now into this process, I share all of the enthusiasm that I ever had about the promise, but it is very difficult to put it in context. It took an IBM team of dozens of people 4 years to win Jeopardy!. Improving our ability to make good oncology decisions is, frankly, much more complex, and it's going to take time. With the knowledge that we have and the competition to make this happen quickly, we are going to see results a lot sooner, but it's not easy.

I urge you to see this as something that will enhance your ability to make good decisions for patients. It will never replace you. The hope is that it would not be a burden to you, but that it would be an advisor to you, just as you would call a trusted colleague about a difficult case or to review what you think is the best decision you can make for your patient. Hopefully, these sorts of computing systems will be able to do the same.

Progress is being made. Stay tuned; you are going to hear more.