Computer-Aided Detection Hinders Mammography

April 6, 2007 — A study shows that computer-aided detection reduces the accuracy of mammography by increasing false-positive results and boosting recall and biopsy rates. Published in the April 5 issue of The New England Journal of Medicine, the work by Joshua J. Fenton, MD, MPH, from the University of California, Davis, in Sacramento, California, and colleagues, already is sparking reaction. In an accompanying editorial, Ferris M. Hall, MD, from the Beth Israel Deaconess Medical Center in Boston, Massachusetts, calls the findings "a substantial hit to this technology." Dr. Hall writes that the study will surprise and disappoint most mammographers.

Initially developed to assist radiologists, computer-aided detection analyzes digitized mammograms and identifies suspicious areas for review by the radiologist. Promising studies led to its approval by the US Food and Drug Administration (FDA) in 1998, and Medicare and many insurance companies now reimburse for its use. Within 3 years of FDA approval, 10% of mammography facilities in the United States adopted the technology and more have done so since. Dr. Fenton told Medscape that his colleagues estimate that number to be about 25% to 30% of facilities today.

"It's not clear just how popular computer-aided detection is," Dr. Fenton said during an interview. "It's still considered a big-ticket item and is seen as a large capital investment in mammography — a not very lucrative area."

Still, Dr. Hall points out that the work is the "most comprehensive analysis of computer-aided detection in breast screening to date." The study involved more than 429,000 mammograms and 2351 cases of cancer that were detected at 43 facilities of the Breast Cancer Surveillance Consortium. During 4 years of observation, 7 of 43 facilities implemented computer-aided detection allowing for a comparison of the performance at these facilities and their individual radiologists before and after the use of computer-aided detection. The facilities that did not implement the technology served as controls.
Not Only Failed to Increase Cancer Detection, but Was Harmful

The investigators found that the use of computer-aided detection not only failed to significantly increase the cancer-detection rate but also was harmful because of the increased number of false-positive results leading to significantly more call-backs and biopsies. These downstream costs, which also may include payments to surgeons and pathologists, account for perhaps one third of the total cost of breast-screening programs, the researchers propose.

"Our study suggests that this technology may not offer a benefit in the way people would have hoped," Dr. Fenton told Medscape. The investigators explain that approximately 157 women would be recalled and 15 women would undergo biopsy to detect 1 additional case of cancer, possibly a ductal carcinoma in situ. After accounting for the additional fees for the use of computer-aided detection and the costs of diagnostic evaluations after recalls resulting from the use of the technology, the group calculates that system-wide use could increase the annual national costs of screening mammography by approximately 18%.

"One possible flaw in the study was the failure to assess the time it takes to adjust to computer-aided detection," Dr. Hall writes in the editorial. "Mammographers initially exposed to computer-aided detection may be unduly influenced by the 3 to 4 marks the software places on each mammogram, with the necessity to ignore the 1000 to 2000 false positive marks for every true positive mark. The adjustment to computer-aided detection has been estimated to take weeks to years."

The researchers also found that computer-aided detection was disproportionately associated with the detection of ductal carcinoma in situ. Dr. Hall notes this is not surprising because computer-aided detection is relatively more sensitive in detecting microcalcifications than in detecting masses. "The relationship of ductal carcinoma in situ to invasive breast cancer remains unclear: all invasive breast cancers probably arise from an in situ monoclonal cancer," Dr. Hall writes, "but many of these lesions may never progress to invasive cancer during a woman's lifetime."

Dr. Hall argues that it took 2 to 3 decades of controversy before it was proved that screening mammography saves lives. "What is the future of breast imaging? I find it hard to believe that we will continue to use mammography to screen up to one quarter of the adult population of the world annually. Mammography is an inherently poor, 2-dimensional projectional method being used to diagnose small, 3-dimensional cancers."

Dr. Hall recommends larger, controlled studies of computer-aided detection that assess not only cancer diagnosis but also the gold standard of mortality. "But," the editorialist notes, "such studies will be expensive, controversial, indeterminate, or quickly passé owing to the emergence of new technology."

N Engl J Med. 2007;356:1399-1409, 1464-1466.