July 21, 2015 - Researchers at the University of Virginia (UVA) Health System have demonstrated that automated volumetric fibroglandular breast density measurement tools are more precise than area-based methods. Results of the study, "Reliability of Automated Breast Density Measurements," recently featured in Radiology, suggest that with lower variability, volumetric breast density is well suited for inclusion in breast cancer risk models. The announcement was made at The Association for Medical Imaging Management (AHRA)'s 43rd annual meeting and exposition, July 19-22 in Las Vegas.
"Breast density is increasingly being considered with other known risk factors to improve risk prediction in order to give women personalized knowledge to make decisions about screening. However, variability in assigned density category may result in changes in recommendations for adjuvant screening. Thus, for consistency, objectivity, and ease of use, breast density measurement ideally should be automated and accurate," said Jennifer Harvey, M.D., professor of radiology at the UVA School of Medicine. "The purpose of this study was to estimate the reliability of area-based methods and automated volumetric breast density measurements using repeated measures."
Thirty women undergoing screening mammography consented to undergo a repeated left craniocaudal examination performed by a second technologist in this prospective study. Breast density was measured by using both area-based and volumetric methods (Volpara software). Discrepancy between the first and second breast density measurements was obtained for each algorithm by subtracting the second measurement from the first and then analyzed with a random-effects model to derive limits of measurement agreement.
Results of the study demonstrate that variability in a repeated measurement of breast density is highest for area-based measurement tools, standard deviation 3.32 percent (2.65-4.44). In contrast, precision was highest for automated volumetric breast density tools, such as VolparaDensity, standard deviation 0.99 percent (0.79-1.33).
"The excellent reproducibility of automated breast density measurements indicates that they would be well suited for inclusion in a breast cancer risk model. This is consistent with results we presented at the San Antonio Breast Cancer Symposium in December that showed that the addition of volumetric breast density improved breast cancer risk discrimination," Harvey added.
For more information: www.uvahealth.com
