Study Finds Computer Software As Effective As Second Radiologist

Main Category: IT / Internet / E-mail
Also Included In: Breast Cancer
Article Date: 06 Oct 2008 - 8:00 PDT

email to a friend   printer friendly   opinions  

Computer-aided detection software for mammograms is nearly as effective as a second radiologist reading the test results to determine the presence of breast tumors, according to a study published on Thursday in the New England Journal of Medicine, Reuters reports (Emery, Reuters, 10/1). For the study -- led by Fiona Gilbert, a radiologist at the University of Aberdeen in Scotland, and sponsored by the British government and Cancer Research UK -- researchers analyzed the results of a previous study involving more than 31,000 British women who underwent routine mammograms. The mammograms in most cases were reviewed once by two radiologists before they were reviewed again by one radiologist and the detection software (Nano, AP/San Francisco Chronicle, 10/2).

Double-reading increases the cancer detection rate and at least 12 European countries use two radiologists to evaluate mammograms. But standard practice in the U.S. is to have one radiologist review the results (Reuters, 10/1). According to the AP/Chronicle, computer-aided detection, or CAD, was developed to help detect more cancers but the added value of CAD continues to be debated. The study found that the detection rate of tumors by two radiologists was 88.7% and the recall rate for further examination was 3.4%, compared with a detection rate of 87.2% and a recall rate of 3.9% by one radiologist and the software (Reuters, 10/1). Researchers found that the software discovered 198 of 227 cancers, compared with 199 discovered by two radiologists.

Gilbert said that CAD could be used to expand screening in Britain and researchers wrote that in places like the U.S. where "single reading is standard practice," CAD has the "potential to improve cancer-detection rates to the level achieved by double reading" (AP/San Francisco Chronicle, 10/2). According to Reuters, researchers did not evaluate the accuracy rate of the new digital mammogram devices and did not consider the potential cost savings of using the software, Reuters reports (Reuters, 10/1).

Reprinted with kind permission from http://www.nationalpartnership.org. You can view the entire Daily Women's Health Policy Report, search the archives, or sign up for email delivery here. The Daily Women's Health Policy Report is a free service of the National Partnership for Women & Families, published by The Advisory Board Company.

© 2008 The Advisory Board Company. All rights reserved.

• Additional
• References
• Citations