The University of Maryland School of Medicine is one of only two universities working with IBM to test the advanced analytics of the company's "Watson" computer for potential health care applications.
The computer program, which may be best known for besting two top champions on the TV show Jeopardy! in February 2011, has shown an amazing ability to comprehend human language, a barrier that has been a challenge for computer designers for many years. Watson can also absorb huge databases and then mine that information quickly.
Eliot Siegel, MD, professor of diagnostic radiology and nuclear medicine, will be leading the project for the School of Medicine.
He says, "The system also has the potential to ingest information from a single patient's electronic medical record in one facility or potentially multiple facilities and also to acquire information from multiple patients. It then has the ability to form multiple hypotheses in a manner similar to the way in which it understands the Jeopardy! question and forms multiple hypotheses."
In the health care setting, the Watson technology may be a powerful tool, helping doctors diagnose patients. Siegel suggests the technology has the potential to result in a renaissance in the application of "artificial intelligence" in medical data mining, data analysis, and decision support.
"I see Watson's capabilities, not as a replacement for physicians, but as an adjunct and tool to organize, highlight, and prioritize information to make a physician more efficient and effective, and improve patient safety," says Siegel. "In a manner similar to a physician who works with residents and fellows and medical students, our physician of the future might utilize this tool to provide improved patient care more cost effectively."
Siegel directs the Maryland Imaging Research Technologies Laboratory at the School. He also is head of imaging at the Baltimore Veterans Affairs Medical Center.
Columbia University is the other institution working on the health care applications for the Watson system.