• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Analysis: Computers No Aid to Mammography

Posted on: Wednesday, 4 April 2007, 18:00 CDT

By ED SUSMAN

The use of computers to assist doctors in reading mammography scans failed to add to the accuracy of screening women for breast cancer, and might have just increased anxiety and costs, doctors said Wednesday.

In a massive, government-sponsored study, researchers said that mammographers who read the scans without computer assistance had fewer recalls of patients for more testing and fewer negative biopsy findings than when the films were re-read by the computer.

In our study, we were unable to show that computer-assisted detection for mammography benefited women, Joshua Fenton, assistant professor of family and community medicine at the University of California, Davis Health System in Sacramento, told United Press International. His study will be published in Thursday's edition of the New England Journal of Medicine.

In an editorial that accompanied the report, Ferris Hall, professor of radiology at Beth Israel Deaconess Medical Center/Harvard Medical School in Boston, said that the study results won't end the use of computer-aided detection but they constitute a substantial hit to the technology.

He said larger studies are needed to examine how and if computer-assisted detection fits into clinical practice.

Every time a computer reads a digital mammography scan, it typically finds two to four areas that need to be rechecked by the radiologist, Fenton said, but only about one mark in 2,000 represents a cancer."

Computer-aided detection identifies suspicious findings on mammograms to assist radiologists, he said. Since the Food and Drug Administration approved the technology in 1998, it has been disseminated into practice, but its effect on the accuracy of interpretation is unclear.

In the study. researchers reviewed data from mammographies performed at 43 facilities in three states, scrutinizing scans of 222,135 women -- a total of 429,345 mammograms, including 2,351 women who received a diagnosis of breast cancer within one year after screening.

Seven facilities implemented computer-aided detection during the study period, Fenton said. When comparing the results before and after implementation, the researchers determined that diagnostic specificity decreased from 90.2 percent before implementation, to 87.2 percent after implementation; the positive predictive value decreased from 4.1 percent to 3.2 percent, and the rate of biopsy increased by 19.7 percent. All those differences reached statistical significance, Fenton said.

The use of the computer increased the cancer-detection rate -- including cases of invasive breast cancers and ductal carcinomas in situ (DCIS) -- from 4.15 cases per 1,000 screening mammograms before implementation, to 4.20 cases per 1,000 screening mammograms after using the computer to assist the readers. That was not a significant difference, he said,

Although the study was performed between 1998 and 2002, Fenton said no studies have been published that indicate there have been significant changes in the ability of computer-assisted programs to improve detection.

This was a real-world comparison of mammographers at community facilities to do these studies, and the results should be generalizable to most practices, he said.

Beth Israel's Hall also noted that researchers are looking at newer imaging devices such as magnetic resonance to conduct mammographies, although those devices also have issues with general use that continually plague screening huge numbers of people, including costs and interpretation.

Source: United Press International

More News in this Category