Could a Computer Out-Diagnose Dr. House? | Smart News | Smithsonian
Current Issue
July / August 2014  magazine cover
Subscribe

Save 81% off the newsstand price!

Keeping you current

Could a Computer Out-Diagnose Dr. House?

Could computers may make diagnostic wizardry a thing of the past?

smithsonian.com

Photo: susansimon

In medicine, the best diagnosticians are held in high esteem by both their colleagues and the public. Like characters from Fox’s hit show, “House,” such superstar physicians possess an encyclopedic familiarity with thousands of maladies and excel at honing in on small clues that less deft experts may ignore. But computers may make diagnostic wizardry a thing of the past.

Such tools are slowly popping up on the market. Jason Maude, a former money manager in London, created Isabel, a program named after his daughter who, at age 3, came down with chickenpox, which veiled a far more serious diseases, necrotizing fasciitis. The flesh-eating infection progressed to the point that, at age 17, Isabel is still having plastic surgery, The New York Times reports. The program named for her aims to eliminate such misdiagnoses and oversights.

The Times also looks at IBM’s Watson for Healthcare, another possible contender in the diagnostic race. Watson can analyze the equivalent of thousands of textbooks every second, whereas the average physician sets aside five hours or less each month to keep on top of the recent medical literature. Watson can understand the nature of a question and review large amonts of information, including a patient’s medical record, textbook and journal articles. While the program does not directly apply to diagnoses yet,  it can offer a list of suggestions to fit a given medical query with a confidence level assigned to each, the Times reports.

Still, while not every physician can be a diagnostic wiz, those who are will likely always have enough work to keep them busy, regardless of whether or not computers enter the clinic. Computers may excel at crunching numbers, but people are naturally good at matching patterns with combined logic and knowledge. Programs like Isabel and Watson may serve as tools for the average physician, but the Dr. Houses of the world will always be in demand.

More from Smithsonian.com:

Telemedicine Predicted in 1925
What Is Killing the Tasmanian Devil? 

 

 

Tags

Comment on this Story

comments powered by Disqus