ABSTRACT
Although mammography screening (Chapter 1) has been shown to contribute to reducing mortality from breast cancer by up to 30% (Tabár et al. 2011), its accuracy can be limited by both the obscuring e›ects of healthy tissue overlying and underlying a lesion or, in some cases, the lack of intrinsic radiographic contrast between malignant and normal œbroglandular tissue. Superposition of tissue results from the projection nature of mammograms where images are formed by transmitted x-rays that must traverse all tissues along a path through the thickness of the breast before detection. Overlap of the more attenuating œbroglandular tissues in particular, can produce complex patterns and regions of considerably increased opacity in a mammogram (Chapter 7). ™is means that image interpretation can be particularly di–cult in breasts with a high proportion of œbroglandular tissue, o¬en referred to as radiographically “dense” breasts (Boyd et al. 2007). ™e sensitivity of cancer detection (fraction of cancers that are detected) has been shown in one study to be only 63% for women with dense breasts, as compared to 87% for fatty breasts (Carney et al. 2003). An estimated 40% of women undergoing mammographic screening (i.e., routine examination of asymptomatic women) have dense breast tissue, so this limitation a›ects a large portion of the screening population (Jackson et al. 1993). Two promising techniques that may overcome the superposition of breast tissue observed in digital mammography are breast tomosynthesis (Chapter 4) and dedicated breast computed tomography (CT) (Chapter 5).
