Breast composition and radiographic breast equivalence

Full text: The composition of the female breast is an allusive variable in screen-film mammography that has a vital role in determining both the radiographic image quality and breast radiation dose. Despite the large volume of research in mammography only two studies known to the author define the chemical composition of the two major breast constituents, fibrous glandular tissue and adipose tissue. Unfortunately there is not complete agreement between these two studies. Upon this basis a variety of breast phantom materials have been suggested and developed. This paper examines the radiographic equivalence of some of these material dealing notably with the formulations given by Hammerstein GR et al (Radiol, 130:485-491, 1979) and ICRU (Report 44, 1989) as well as BR12 (White DR, et al, Brit. J. Radiol. 50:814821, 1977), water and lucite. Radiographic equivalence was calculated using a computer simulation of mammographic spectra (after the example of Birch and Marshall (Phys Med Biol 24(3):505-517, 1979) and Tucker et al (Med. Phys. 18(3):402-407, 1991) and fitted to available FDA data) applied to attenuation data of phantom materials in turn based on elemental linear attenuation coefficient data. A software algorithm was written that iteratively calculates the thickness of a nominated material required to match the attenuation of another material of particular thickness. The equivalence of 2 - 8 cm of lucite to various phantom materials was calculated for molybdenum target, molybdenum filter (Mo/Mo) x-ray beams from 25 to 35 kVp and for Mo/Rh and Rh/Rh beams from 27 to 35 kVp. It was found that for Mo/Mo beams the equivalent thickness of breast phantom material for certain thicknesses of lucite was largely independent of x-ray beam energy with the range of thickness equivalences being less than 3% at 8 cm of lucite and less than 0.5% at 2 cm of lucite. The RMI 156 breast phantom was found to be equivalent to 42 cm of Hammerstein 50% glandular/ 50% adipose phantom material for Mo/Mo beams at 27 kVp, increasing to 43 cm for the range of Rh/Rh beams. While some question does remain on the accuracy of the present composition of breast constituents it is clear that various breast compositions can be successfully simulated by using an appropriate thickness of lucite material. The use of lucite breast phantoms is then seen as an effective and reliable way of evaluating mammographic image quality and for calculating breast dose for varying required breast compositions