In theory, one could investigate calcification detection by organizing clinical trials to measure detection for different image qualities. In practice, such trials would be very expensive due to the low incidence of cancer in screened populations and there would be ethical issues if there were multiple radiation exposures on the same women. The alternative approach used here mimics mammographic imaging by inserting simulated calcification clusters into normal digital mammograms acquired with an amorphous selenium DR detector. The images were then modified using a method described previously to have the appearance of images acquired on a CR detector typical of the single-sided granular type widely used in Europe. The images were also modified to have the appearance as if they were acquired on both systems at lower dose levels. An advantage of this approach is that the inserted calcifications and anatomical backgrounds were identical between the different test sets used in the observer study. This is important because variations in anatomical background are expected to have a major impact on cancer detection and are a confounding factor in studies of this type. Thus, our approach allowed paired comparisons of images of different image qualities and was suitable for FROC analysis, overcoming the limitation of some of the previous studies described above.