The article on computer-aided detection (CAD) by Fenton et al in the April 5, 2007 issue of the New England Journal of Medicine raises serious questions about the utility of CAD in the interpretation of mammograms. Several points should be considered.
Reported decreased accuracy of mammography when using CAD was due to increased false positives, not to fewer cancers being detected. Refinements and improvements in CAD software are continually being made and it is possible that newer versions of CAD systems have better performance parameters. The study was designed in such a way that it was impossible to determine how many cancers in their series would have been missed without the use of CAD.
Finally, there was indeed a statistically significant increase in the detection of ductal carcinoma in situ with CAD. The debate about whether or not this is a desirable outcome of screening mammography continues, but the value of improved detection of DCIS should not be discounted.
Several studies with different designs have demonstrated incremental detection of cancers with the use of CAD. All have shown a trade-off with increased false positives, although to a modest extent. Other studies have shown that using CAD is comparable to double reading by two radiologists.
Clearly, CAD is not a substitute for human interpretation, and those using this tool must use it properly, as an adjunct to, rather than a replacement for, careful mammographic evaluation. Vendors also should strive to continue to improve the performance of CAD systems.
Meanwhile, there is a fair amount of evidence outside the current study to suggest that, when properly used, CAD may indeed be worthwhile and it seems unwarranted to consider abandoning this potentially valuable tool at this time.