CAD algorithms are written to identify specific cancer morphologies. It can minimize mistakes in interpretation when radiologists tire from reviewing mammograms repetitively. The advantage of CAD is that it flags areas of concern for a radiologist to reexamine. Tommy E. Cupples, MD, a private practice radiologist at ImageCare in Columbia, SC, USA, has been a leading proponent of CAD. According to Cupples, CAD can direct the radiologist’s attention to possible areas of concern, but it cannot establish the diagnosis.


Valuable in Flagging Small, Early-Stage Lesions, and Lesions
“It still remains up to the radiologist to determine whether or not the CAD finding is ‘actionable,‘” Cupples says. And in that sense, he explains, the technology is only as good as the accuracy of the mammographer. Where he has found CAD to be especially valuable is in flagging small, early-stage lesions, and lesions in younger women. “These are the ones that are most likely to be overlooked and where CAD makes the most difference,” says Cupples.