A comparison of fixed and variable kVp technique protocols for film-screen mammography

Mammographic image quality, contrast and dose for a variable tube potential (kVp) technique protocol for film-screen mammography have been investigated. In this protocol, the tube potential is increased for larger breast thicknesses. Comparisons were made with fixed kVp protocols, in which the tube potential is kept constant and the breast thickness compensated for by prolonging the exposure ("fixed kVp" protocol). All measurements were performed on a mammography unit with a molybdenum target and filter. Image quality was quantified by image contrast, image detail detection and the minimum detectable dimension of low contrast objects. It was demonstrated that for a compressed breast thickness of less than about 40 mm, varying the tube potential had a negligible effect upon dose but a significant effect upon image quality. For a compressed breast thickness greater than about 60 mm, the effect of the tube potential upon image quality was much reduced; however, the effect upon dose was significantly greater. The variable kVp protocol takes advantage of this feature to yield a significantly lower dose for thicker breasts with a small reduction in image quality, often only within experimental uncertainty. For an exposure under automatic exposure control, increasing the tube potential from 26 kVp to 30 kVp for a breast of a reference tissue composition (50% adipose and 50% glandular) with a compressed thickness of 60 mm reduced the mean glandular dose from 6 mGy to 3.9 mGy (-35%), but increased the minimum detectable dimension of a low contrast mass from 0.8 (+/- 0.1) mm to 1.1 (+/- 0.1) mm. Adopting a variable kVp protocol led to a median patient mean glandular dose per film of 2.7 mGy, nearly independent of compressed breast thickness. In our survey, the mean age of women presenting for mammography is younger and the mean compressed breast thickness is less than reported from screening centres. This suggests that there will be a higher proportion of denser, glandular tissue in the breasts incorporated within this survey than for surveys from screening centres. The clinical use of the variable kVp protocol allows the extraction from patient data of separate changes in breast composition which are due to patient age and breast thickness. It is concluded that the reference breast tissue composition is not an accurate representation of the women presenting at this centre.