The Jury is out on CAD for Mammography
“Computer-aided detection,” with a name like that, what’s not to love? Everyone can use a helping hand from time-to-time, right? Of course, but is CAD the best kind of help? Opinions among Mammography experts differ.
The Product
Computer-aided detection (CAD) involves the use of software and computers to bring suspicious areas on a mammogram to the radiologist’s attention. The image is displayed on a monitor after the radiologist has done the initial review of the mammogram with “suspicious” areas highlighted for further study.
Opposing Views
The Radiological Society of North America maintains that CAD systems can help radiologists diagnose more early-stage cancers than mammograms alone and there are many radiologists who agree. (RSNA 2012) Others, however, question how many cancers CAD software can identify correctly. They hold that the CAD device is not as effective as proponents believe due to the significant number of false positive CAD marks that can occur on a screening. False positives, at best, add extra time and frustration to the work of a physician who may be reading 70+ mammograms per day. At worst, they cause patients to be recalled for unnecessary testing. (Doheny 2011)
It seems that a significant reduction in false positive rates is the only way for CAD systems to silence their detractors. It is only when these rates approach those of a radiologist that CAD will be fully embraced as a second opinion tool by the industry at large.
The Future
There is new CAD technology nearing release, pending approval from the FDA. This new generation of CAD will serve as the next step in convincing more radiologists to embrace it with open arms. Whether or not this will happen on a grand scale has yet to be seen, but mammogram providers on either side of the debate can take comfort in the ongoing research to develop increasingly accurate tools to help them be more successful in saving lives.
...............
Meet the authors: Jason Block and Jordan Batterbee
Browse related blog articles by category