Mammography is uniquely effective in early detection of breast cancer; however, because this procedure is used for screening asymptomatic women and breast tissue is sensitive to radiation carcinogenesis it is important to monitor the radiation dose delivered to the breast.7 Although there is ample evidence that the resulting benefits will substantially exceed potential risks (Section 7), the conservatively safe assumption of a linear non-threshold dose-risk relationship requires that the dose be minimized while necessary image quality is maintained. An early study conducted by the Center for Devices and Radiological Health correlated faulty technique with excessively low, as well as excessively high doses (Jans et al, 1979). A recent study discussed the relationship between phantom failure rates and radiation dose in mam-mography accreditation (Haus et al., 2001). Since patient dose levels can, therefore, provide a useful check on the diagnostic adequacy of mammography technique as applied in practice, dose values are useful for assessing and monitoring risk, selecting techniques, and verifying their proper clinical application.
There should be a suitable dose index for each radiographic procedure to help in selecting among alternatives. In addition, dose values should characterize the probable risk of radiation carcino-genesis in the female population studied. Finally, to be practical, the dose-evaluation procedure must be easily applied in a clinical setting. Investigators did not always consider all of these requirements; for this reason, the literature has shown a significant range of doses for the same techniques. Previous work has been carefully reviewed in preparation of this Report in order to determine a preferred approach to dose evaluation.
7The term "dose" in this Report is used generically when not referring to a specific quantity, such as "mean glandular dose."
Section 5.2 explains why the mean glandular dose most appropriately characterizes radiation risk from mammography and discusses several other dose quantities that have been widely used. A brief review is also included of x-ray exposure, air kerma, and absorbed dose concepts.
Section 5.3 discusses the basic approach of the recommended method and explains how mean glandular dose is determined. Section 5.4 presents a summary of current dose recommendations, doses measured in nationwide surveys, and also the relationship of phantom dose determinations to actual dose delivered in clinical examinations. Section 5.5 summarizes the recommended doseevaluation procedure.
5.2 Risk-Related Dose
Three considerations must be kept in mind in estimating the potential carcinogenic radiation risk of mammography. First, glandular tissue is the most vulnerable in the breast as compared to adipose, skin and areolar (nipple) tissues (Hammerstein et al., 1979). (In this context "glandular tissue," which includes acinar and ductal epithelium and associated stroma is assumed to have equal sensitivity throughout.) Second, the mean rather than maximum dose to the glandular tissue most usefully characterizes risk of carcinogenesis and is consistent with an assumed linear dose-response relationship. Third, the population of primary interest is women 40 y and older since younger women are likely to have only diagnostic mammographic examinations because of physical findings (or a single baseline screening study); it is therefore, reasonable to assume that the dose calculations apply primarily to breasts containing a larger fraction of adipose tissue found primarily in older women (Section 5.3.1).
The major variables that affect the breast dose per view delivered in a mammographic examination are: the choice of image receptor, the x-ray beam energy (HVL and kilovolt peak), the degree of breast compression, and the breast size and adiposity (Table 5.1). Xeroradiography systems are not considered since their
Table 5.1—Variables affecting breast dose.
• Breast size and adiposity
- Thickness: Exerts great effect.
- Field size: Exerts minimal effect.
- Adiposity: Only a moderate effect.
- Breast dose is reduced when beam energy is increased. This may be at the cost of reduced image contrast.
- There is a slight variation of breast dose for constant HVL, but varying tube potential.
- The use of rhodium filters or targets can reduce breast dose for thicker or more glandular breasts. This may be at the cost of reduced image contrast.
• Types of image receptor
- Optimum beam HVL for screen-film technique is 0.3 to 0.4 mm Al.
- Breast dose is determined by required optical density.
- Optimum settings for digital systems are individually determined for each type of system. Breast dose is determined by required SNR.
• Grid versus nongrid
- The bucky factor for most mammography grids is about 2 to 2.5 leading to an increase in breast dose by this factor for grid versus nongrid techniques.
• Degree of breast compression
- Firm compression reduces breast dose as much as 50 percent and is indispensable for best image quality with all techniques.
manufacture has been discontinued. It should be noted here that the dose associated with any digital technique is not governed by the need to obtain a proper image density or brightness, since that may be obtained for almost any dose with appropriate manipulation of the window and level controls. Doses for those techniques are normally determined by the need to obtain an acceptable SNR in the image.
Increasing the x-ray beam energy tends to reduce breast dose but at the cost of image contrast. As a result, there is an optimum and quite narrow range of beam quality for screen-film breast imaging. In recent years, special target-filter combinations, such as Rh/Rh or Mo/Rh, have been introduced to obtain high-quality images of breasts of greater thickness or with a higher percentage of glandular tissue.
Firm compression greatly reduces the dose (Figure 5.1). The effect of firm compression is to spread the breast volume laterally, thereby significantly reducing the x-ray path through the breast. Possible dose reductions as a result of breast compression can exceed 50 percent for screen-film techniques. Compression also greatly modifies the breast shape. By making the sagittal and transverse cross-sections of the breast and its glandular component more nearly rectangular (Figures 5.2a and 5.2b), the compression simplifies the geometric configuration of breast structures and permits use of the computational model of Figure 5.2c in dose determinations (Section 5.3).
Female breasts vary greatly in size and adiposity resulting in a significant range of dose values for a given technique. The compressed breast thickness affects dose to a great degree and, hence, must be specified to obtain accurate dose values. Although the breast area when compressed also varies greatly, the effect on dose is relatively small. For example, an increase from 35 to 270 cm2 changes the dose to the breast by <10 percent (Dance, 1980). Finally, a breast containing a high fraction of adipose tissue is more readily penetrated than one containing a high fraction of fibroglan-dular tissue, and thus, a fatty breast receives a lower dose per view from the same technique. However, simple dose-evaluation procedures can yield mean glandular dose values reasonably independent of moderate differences in breast composition so that simple corrections can be applied (Stanton et al., 1984).
Was this article helpful?