Gisella Gennaro PhD, Presenter: Nothing to Disclose

Henri Souchay, Abstract Co-Author: Nothing to Disclose

Cosimo di Maggio MD, Abstract Co-Author: Nothing to Disclose

Luc Katz MS, Abstract Co-Author: Nothing to Disclose

Manuela La Grassa MD, Abstract Co-Author: Nothing to Disclose

Luigi Pescarini MD, Abstract Co-Author: Nothing to Disclose

Claudio Alberelli PhD, Abstract Co-Author: Nothing to Disclose

et al, Abstract Co-Author: Nothing to Disclose

Evaluate the impact of dose reduction on diagnostic performance in a population undergoing full-field digital mammography.

A clinical sample of 305 cases was collected, including 56 cancers, 150 benign lesions and 100 with no lesions; 111 cases were found with microcalcifications, 81 with masses and 13 with both masses and calcifications. Each patient had two images of the same breast acquired with a full-field digital mammography unit (GE Senographe 2000D): one image at standard dose and another reducing dose by a factor of 30% to 50%, based on a previous phantom study. Three experienced and independent radiologists using soft-copy reading recorded finding localization, type, confidence of presence and rated the probability of cancer using an "extended” BIRADS 7-step scale. Lesion classifications from each radiologist were compared with the truth (histology or 2-years follow-up) and ROC analysis at standard and reduced dose was performed.

At 95% confidence level, difference of areas under ROC curves at standard and reduced dose was not statistically significant. ROC curve area ranges for standard and reduced dose images overlapped for each and all radiologists. The overall ROC curve area for standard dose was 0.811, with the 95% confidence interval 0.756-0.893, inclusive of the ROC curve area value for reduced dose (0.796). Variations in lesion assessment among different observers were dominant, independently of dose level.

Inter-observer variability on lesion classification was greater than the effect of dose reduction. Results support the opportunity of reducing dose in digital mammography without significant impact on clinical decision.

H.S.,L.K.,C.A.: Are employees of GE Healthcare

Gennaro, G, Souchay, H, di Maggio, C, Katz, L, La Grassa, M, Pescarini, L, Alberelli, C, et al, , Dose Reduction in Digital Mammography: Impact on Diagnostic Performance.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4417315.html