Physical Aspects of Breast Radiography

Abstract

Because of renewed and widened interest in breast radiography, many additional studies are contemplated to evaluate its clinical potentialities and limitations. It seemed to us that optimum technical procedures are essential to carry out any such program. Previous work on selection of the best technical factors has been largely empirical, with little assistance from those interested in the basic physics of the problem. This has resulted in a variety of technics, difficult to evaluate. For example, those used by two of the best-known investigators in mammography (Table I) differ in every detail! In this paper an attempt is made to provide a physical basis for the selection of acceptable mammography technics. Our approach includes the following: I. Production of a satisfactory breast roentgenogram—analysis II. Studies using a phantom breast6 III. Dosimetry IV. Discussion—technics and practical considerations. I. Production of a Satisfactory Breast Roentgenogram—Analysis Pathognomonic signs of malignant growths of the breast include the following (1–4): 1. Small calcific deposits, in clusters of irregular distribution: Calcium occurs in specks so tiny (frequently less than 1 mm.) that they are clinically observed in only 30 to 50 per cent of proved malignant tumors (2, 3). However, they are reported in up to 70 per cent of proved surgical specimens (5). Unlike soft-tissue lesions, calcium may be detected even in young and cystic breasts. 2. Soft-tissue lesions, both carcinoma and marginally situated fibrous tissue (4): These produce adequate subject contrast primarily in fatty breasts. Three technical requirements must be met if these signs are to be dependably recognized in a roentgenogram. First, very fine detail must be present for detection of calcium deposits and structural detail in and about soft-tissue lesions. Second, high contrast is needed to highlight both small calcium specks and soft-tissue lesions. Finally, exposure times should be preferably three seconds or less, as patient motion blur can destroy an otherwise good roentgenogram. Detection of a Roentgenographic Shadow In all x-ray procedures, recognition of disease involves both detection of significant signs and correlation of those signs with a previous memory pattern (6, 7). The importance of training and specific experience in the second aspect of this process cannot be overemphasized. However, even the most experienced radiologist must first notice significant details. This process of detection involves the following considerations: 1. Optimum film brightness.7 Optimum film brightness is necessary to assure adequate retinal illumination. 2. Adequate size of the image on the film: Separation of the foveal cones sets an absolute limit of about one minute of arc in detection of small objects (8). At a viewing distance of 25 cm. this corresponds to an image on the film of less than 0.1 mm.