More Content - Including Podcasts

Tuesday, June 19, 2012

IBM and Evidence Based Medicine Decision Support Systems

Jeffrey Betts from IBM makes a repeat appearance at the conference to discuss how the IBM Watson project can be leveraged for evidence based medicine and clinical decision support systems.

The Watson systems understands natural language, generates and evaluates hypothesis, and learns by homing its own decision algorithms. The solution has been developed hardware agnostic, but is generally run on parallel HPC systems for optimal response times.

The key point of this lecture is that computers are by default and historically poor at responding to unstructured data. Human minds have natural abilities to view unstructured data and identify patterns, and this is the goal of a true expert system such as Watson aspires to be.

Jeffrey takes us through screen shots of a case of an oncologist using Watson to assist with a consult. It was an interesting update for which no one in the audience had any questions.

- Posted using BlogPress from my iPad

Location:Hwy 97 S,Kelowna,Canada

No comments: