Potential for lower doses in digital mammograph Tumor detection in hybrid images using a dose reduction simulation method

Abstract: Purpose: To determine how image quality linked to tumor detection is affected by reducing the absorbed dose to 50% and 30% of standard dose levels, i.e. an average glandular dose of 1.26 mGy for a standard breast according to European guidelines. Material and methods: Tumors were computer simulated and inserted into 40 out of 90 normal, unprocessed images acquired from the screening department at Malmö University Hospital using a Siemens Mammomat Novation full-field digital mammography unit. All 90 images were then dose-reduced by adding simulated quantum noise to represent images acquired at 50% and 30% of the original dose levels set by the automatic exposure control (AEC). This yielded 270 images that were subsequently processed for final display. One radiologist and 3 physicists participated in this study in which they searched for and marked the positions of the tumors and the degree of suspicion. The analysis of results were carried out using the jackknife free-response ROC (JAFROC) method together with analysis of variance (ANOVA).Results: In the tumor detection study, the cumulative figure of merits (FOM’s) calculated from JAFROC scoring were 0.61, 0.66 and 0.67 for the 100%, 50%, and 30% dose levels, respectively. ANOVA revealed an F-stat and p-value of 1.90 and 0.15, respectively, indicating no statistical difference in tumor detection between any two pairs of scores.Conclusion: For the simulated tumors used in this experiment, there was no significant change in detection by increasing quantum noise, indicating a potential for dose reduction. Further studies could involve using different shaped masses or microcalcifications, different processing algorithms and different window/level settings.