Basic physics and doubts about relationship between mammographically determined tissue density and breast cancer risk

Numerous studies have suggested a link between breast tissue patterns, as defined with mammography, and risk for breast cancer. There may be a relationship, but the author believes all of these studies have methodological flaws. It is impossible, with the parameters used in these studies, to accurately measure the percentage of tissues by volume when two-dimensional x-ray mammographic images are used. Without exposure values, half-value layer information, and knowledge of the compressed thickness of the breast, an accurate volume of tissue cannot be calculated. The great variability in positioning the breast for a mammogram is also an uncontrollable factor in measuring tissue density. Computerized segmentation algorithms can accurately assess the percentage of the x-ray image that is "dense," but this does not accurately measure the true volume of tissue. Since the percentage of dense tissue is ultimately measured in relation to the complete volume of the breast, defining the true boundaries of the breast is also a problem. Studies that purport to show small percentage differences between groups are likely inaccurate. Future investigations need to use three-dimensional information.