The effect of change in dose on microcalcification detection has also been investigated in the past. Two studies using phantom and clinical images with simulated calcifications found a significant reduction in detection at quarter the dose level but not half the dose level. Both these studies used a location-known exactly experimental paradigm. Using this paradigm is a limitation if it is aimed to predict the clinical performance that requires both search and classi- fication. A further study inserted simulated mass and calcification clusters into clinical images. These images were then simulated to 50% and 30% of the original dose level. The entire image was viewed during the study and observers were required to localize the lesions prior to rating the lesion on how confident they were that they had identified a lesion. This study found a significant reduction in detection at 50% and 30% the original dose level.