Computer-assisted decision support systems, called CAD, have been used for a number of years to highlight areas of potential abnormality in radiological examinations and today there's no question of their value, says Paul Chang, M.D., a professor and vice chair of radiology informatics at the University of Chicago School of Medicine.

During a three-part educational session at the Radiology Society of North America's annual conference, Nov. 27-Dec. 2 in Chicago, Chang will discuss integrating CAD into existing workflows. Other parts of the session will cover a short history of CAD and changes to the practice of medicine from an evolution in the understanding of the molecular basis of disease.

But Chang also will advocate expanding the concept of CAD beyond detecting abnormalities, so that decision support becomes a "true partner" for clinicians. CAD systems, he notes, often are tuned to certain sensitivities and they tend to maximize the sensitivities.

For instance, young radiologists tend to "overcall" abnormalities--to see more than what is really there--and using a CAD can encourage this. "We need instead of having CADs frozen in one sensitivity to make them tunable to physicians' strengths and weaknesses," Chang says. Frozen sensitivity results in more false positives if a radiologist over-calls, and more false negatives if a radiologist is conservative. To prevent this, over-callers and under-callers are regularly paired in Britain to read the same exams, he says.

CAD systems, however, can be trained by tuning to be an over-caller or under-caller by being opposite of the physician reading the exam, and it's time go that route, Chang believes.

It's also time to get CAD more seamlessly integrated into radiology workflows to optimize image flow, he adds. "It's no good if CAD is inefficient, especially as reimbursement drops about 35 percent."

Course No. RC523, "Minicourse: Computer-Assisted Decision Systems in Radiology--The Hope, the Hype and the Hard Truth," is scheduled at 8:30 a.m. on Wednesday, Nov.  30, in Room S504AB in McCormick Place. More information is available at rsna.org.

 

Register or login for access to this item and much more

All Health Data Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access