Study Assesses Reproducibility of Computer-Aided Detection
In digital mammograms, reproducibility higher for true-positive CAD marks than false-positive
THURSDAY, Jan. 3 (HealthDay News) -- Practitioners who are interpreting mammograms using computer-aided detection (CAD) should consider breast cancer when the system places CAD marks in the same breast area in initial and follow-up mammograms, given the reproducibility of true-positive marks, according to research published in the January issue of Radiology.
Seung Ja Kim, M.D., of Konkuk University Hospital in Seoul, Korea, and colleagues evaluated CAD system performance using 93 women with known breast cancer who had initial and follow-up digital mammograms taken less than two months apart.
The sensitivity of the CAD system for detecting masses was 91 percent at the initial digital mammography and 89 percent at follow-up. For microcalcifications, sensitivity was 100 percent at both sessions. False-positive mark rate was 0.29 per image at the initial session, and 0.27 for follow-up. The reproducibility of CAD marks was 80 percent for true-positive masses, 92 percent for true-positive microcalcifications, 9 percent for false-positive masses, and 8 percent for false-positive microcalcifications, according to the authors.
"CAD performance may depend on background breast density. In our study, mass sensitivity was significantly higher for fatty breasts than for dense breasts, whereas microcalcification sensitivities were unaffected by breast density," the authors write. "The early detection of breast cancer in the absence of microcalcifications, particularly in dense breasts, appears to be a difficult task, both for radiologists and for CAD systems. Several CAD algorithms used for detection of masses on digital mammograms have been reported to offer improved detection rates," they add.