The current standard of care for breast cancer screening is x-ray mammography for women over the age of 40. A technician that compresses the breast and takes pictures from different angles, creating a set of images of each breast, usually performs this technique. In the set of images, called a mammogram, breast tissue appears white and opaque, while fatty tissue appears darker and translucent. X-rays travel unimpeded through soft tissues; however, cancerous tissue absorbs x-rays and can show up on the film as white areas. In a screening mammogram, the breast is x-rayed from center to side.
However, a diagnostic mammogram focuses in on a particular lump or area of abnormal tissue. This examination usually takes about 30 minutes. Yearly screening mammography results in sensitivity (proportion tests that correctly indicate a woman has cancer) ranging from 71 to 96 percent and specificity (proportion of tests that correctly indicate that a woman does not have cancer) ranging from 94 to 97 percent.18 However, several factors influence the correct detection of breast cancer, such as age, breast density, hormone replacement therapy, image quality, and experience of the radiologist.18
CAD involves the use of computers to identify suspicious areas on a mammogram after the radiologist’s initial review of the mammogram. CAD double-checks the work of the radiologist to help avoid possible oversights. In 1998, the FDA approved the first CAD system, ImageCheckerTM (R2 Technology, Inc., Los Altos, CA). This device can either scan a mammographic film with a laser beam and convert it into a digital image, or obtain images directly from a digital mammography system. The radiologist can see if any of the highlighted areas were missed on the initial review and require further evaluation.
Initial studies show CAD technology may improve the accuracy of screening mammography by reducing the number of missed cancers.3,13 A 2004 study reported that the use of CAD was not associated with statistically significant changes in recall or breast cancer detection rates.15 However, all radiologists in that study were considered breast imaging specialists, and the results of this study should not be extrapolated to use by community radiologists who vary widely in their proficiency.11 The greatest clinical value of CAD probably does not lie in its ability to raise the performance level of all breast imagers, but rather in its potential to bring the performance level of general radiologists to that of breast imaging specialists.35