The Women's Health Resource. On the web since 1997.

FDA Approves Computer-Aided Detection of Breast Cancer on Mammograms (dateline June 3, 2003)


The U.S. Food and Drug Administration recently approved the use of new computer-aided detection (CAD) MammoReader software made by ICAD, Inc., which helps doctors detect breast abnormalities on mammogram films. In January 2003, ISSI, Inc. received FDA approval for their CAD mammography system. CAD works by digitizing a mammogram film and then highlighting areas that may be represent suspicious breast abnormalities. Radiologists can decide whether the computer-highlighted areas require additional evaluation, based on the patient’s situation. Similar software made by R2 Technology, Inc. is also FDA-approved and on the market. These recent government approvals suggest that radiologists may increasingly rely on computer-assisted technology to help diagnose breast cancer.

Computer-aided detection (CAD) technology is a recent advance in the field of breast imaging. The CAD technology essentially works as a second pair of eyes, reviewing a patient's mammogram film after the radiologist has already made an initial interpretation. If the computer software detects any breast abnormalities or "regions of interest" on the mammogram film, it marks them. The radiologist can then go back and review the mammogram film again to determine whether the marked areas are suspicious and require further examination (with additional imaging tests or biopsy). With the CAD technology, the radiologist still makes the final interpretation of the mammogram.

Using sophisticated pattern recognition computer software, the CAD technology is designed to detect the following abnormalities on mammogram films:

  • Patterns of bright spots that suggest microcalcifications (tiny calcium deposits that may indicate cancer)
  • Dense regions with or without radiating lines that suggest breast masses or distortions

A study presented at the 2000 Radiological Society of North America (RSNA) meeting found that using R2 Technology’s CAD software increased the detection of breast cancer by approximately 20%. Among the 12,860 women in the study, 49 cancers were detected:

  • 32 cancers were detected by both CAD technology and by the radiologist, who interpreted the women’s regular mammogram films
  • nine cancers were detected by the radiologist alone
  • eight cancers were detected by the CAD technology alone (the radiologist then reviewed the mammogram films again to find the suspicious areas)

Referring to his company’s MammoReader CAD system, Michael Palazzola, President of Instrumentarium Imaging Inc., said in a news release that, "This is an important breakthrough in the fight against breast cancer. He continued, "Its pricing, starting at $99,500, allows all breast care centers the opportunity to offer this product, thus benefiting a larger population who otherwise would not have access to this technology."

In a study reported by ISSI, researchers evaluated 327 cancer cases and found that 23% of the women could have been diagnosed with breast cancer at an average of 14 months earlier than they had been if their mammograms had been read with the assistance of the MammoReader CAD system. Similarly, ICAD boosts that their new CAD MammoReader software has a 90.5% sensitivity for detecting breast masses.

Though it appears as though the use of computer-aided detection of breast abnormalities on mammogram films will increase, some radiologists worry about potential limitations of the technology. For example, the CAD technology can mark "normal" areas on mammograms as abnormalities, which may lead to the ordering of additional unnecessary and costly breast imaging and/or biopsies.

Nevertheless, computer-aided detection has potential to help detect breast cancer in earlier stages, when the chances of surviving the disease are the greatest. Standard film mammography, the currently the gold standard of breast cancer detection, detects approximately 85% of all breast cancers. Computer-aided technologies have the potential to increase that detection rate.

Additional Resources and References