Advanced clinical decision support with IBM’s Watson
Eliot L. Siegel, M.D. (invited), Professor of Diagnostic Radiology and Nuclear Medicine, School of Medicine, University of Maryland
Eliot calls out SIRI in new iPhone – AI for the common man.
We are just beginning to explore the potential for Artificial Intelligence in Health Care.
IBM software uses Deep Q/A. A set of algorithms that uses parallel processing to discover data. It is not a database of Questions and Answers.
Crowd sourcing with multiple computer algorithms and a rating system to determine best possibility. Also weighing positive and negative evidence.
Why do we need this in health care:
Diagnostic errors outnumber other medical errors by 2-4 times.
Clinicians jump to conclusions and short circuit the diagnostic process.
Personalized medicine is emerging and this adds many other factors to the diagnostic process. AI can help with this analysis.
To improve Dr. Watson’s effectiveness we need to bring book knowledge, experiential learning and access to EMR data.
The challenge is that there is plenty of data out there but they are not public and not discoverable.
We also have to teach Dr. Watson bedside manners.
There is a lot of potential – it is like a promising medical student.