Quality assurance in screening mammography

This evaluation was a cooperative effort between ECRI and the Swedish Testing Institute for Medical Supplies (Sprima AB), Stockholm. As noted in our last mammography evaluation, we are initially focusing on mammography units in our series of radiology and imaging studies because (1) improvements in technology (e.g., lower radiation doses) now make routine screening feasible; (2) third-party payers are supporting screening programs, and (3) breast cancer is a major threat to women's health. Again ECRI will provide physicists and radiologists with our raw test data on request. In our previous evaluation, we discussed the role of mammography in reducing breast cancer mortality rates and tested the units for screen-film mammography and xeromammography capability, basing our ratings primarily on the units' ability to safely and consistently produce acceptable images with minimal patient dose during screen-film mammography. Below, we focus on the definition, measurement, and significance of image quality within the context of screen-film mammography. We evaluated seven mammography units from seven manufacturers. We also retested the Automatic Exposure Control (AEC) mode of the Soredex Mamex dc Mag, which we evaluated last year and rated Conditionally Acceptable because of its AEC performance (see "Soredex Mamex dc Mag"). Our ratings are based on our minimum criteria of acceptability: that the units safely and reliably produce the best possible image quality while delivering the lowest possible dose to the patient. Except for the Kramex HF-45, all of the evaluated units-the Acoma ESP-200, GE-CGR Senographe 600T, Instrumentarium/Ausonics (I/A) Alpha III, Lorad M II-D, Philips MammoDiagnost UC, Picker Sureview, and the Soredex Mamex dc Mag-exceed our minimum criteria and are rated Acceptable for both screening and magnification (diagnostic) applications. The Kramex HF-45, the only dedicated screening unit we evaluated, is rated Acceptable-Not Recommended because of its limitations in achieving optimal image quality. The units are ranked according to their AEC performance and human factors (ergonomic) design, as well as on technical phantom image quality results, within the two applications. Clinical phantom image quality results are listed, but were not used to rank the units. Refer to our previous study for further discussions of breast imaging technology and terminology, as well as radiation risks. Terms set below in small caps are defined in the Glossary.