Mammographic determination of breast volume: comparing different methods

The purpose of this study was to compare the accuracy and reproducibility of different methods for calculating breast volume when using measurements made on mammograms. The volumes of 32 breasts were determined by pathologic evaluation of mastectomy specimens. Two radiologists independently measured breast height and width on the preoperative craniocaudal mammograms and measured height, width, and width at half-height on mediolateral oblique mammograms. Compression thicknesses used on the craniocaudal and mediolateral oblique projections were recorded. Volume was then calculated using six different formulas. The accuracy of each method was determined and compared using bivariate and univariate linear regression analyses. Interobserver variability in measurement was also assessed. The most accurate method for calculating breast volume was the one that assumed a half-elliptic cylinder shape for the compressed breast in the craniocaudal projection. Measurements made on the craniocaudal view were more reproducible than those made on the mediolateral oblique view. Breast volume can be accurately and reproducibly determined on mammograms by making two measurements on the craniocaudal view and knowing the compression thickness. This information may be useful to plastic surgeons, investigators who study parenchymal patterns, and physicians who examine cancer patients being considered for breast conservation surgery.