Fluke Biomedical 18-222 - Manual (Page 5)

1-1
Section 1
Introduction
1.1 Background
The American Cancer Society and American College of Radiology guidelines for the screening of asymptomatic
women have made over 50 million women candidates for mammography. In view of the staggering numbers
involved, it is critically important that simple but reliable methods be developed to assess system performance, to
assure consistent system performance, and to assure consistent production of diagnostically useful images (1, 17).
1.2 The Development of the Standard of Reference
Phantoms for use in mammography should simulate a real breast as closely as possible (2). A list of desirable features
for such a phantom can be found in Section 2, page 2-2. Note that the phantom should be able to test for both
image quality and dose if system performance is to be evaluated. The phantoms must also be easy to use and yield
images that may be unambiguously interpreted.
In developing the tissue-equivalent/realistically shaped phantom:
• Image Contrast may be measured quantitatively with standard densitometers though the use of the
embedded step wedge.
• Dose may be calculated by "TLD" or by ion chamber placed on top of the phantom and converted to average
glandular dose through conversion tables (3.6 and 3.7) in NCRP Report #985 (2). A suggested dose chart is
shown in Figure 2-1.
• Resolution - Simulated tumors and microcalcifications of known size and location are embedded in the
phantom for qualitative evaluation. The smallest microcalcifications and tumors are small enough that they
will not normally be detected.
1.3 The Realistically Shaped, Tissue-Equivalent Series of Breast
Phantoms
Shape
Standard dental modeling techniques were used to obtain molds of the compressed right breast of a volunteer
female subject. This breast is 4.5 cm thick and 18 cm in width.
Materials
Tissue-equivalent resin molding techniques were used. The system of resins used have been developed over the past
six years to permit mimicking of any body tissue at different diagnostic x-ray levels. The elemental composition of the
simulating tissue as compared to Hammerstein's analysis (11) of human tissue is shown in Table 2-1. Also shown in
Table 2-2 are comparisons of linear attenuation coefficients for actual and simulated tissue.