Investigation of mean glandular dose versus compressed breast thickness relationship for mammography

• Published in: Radiation protection dosimetry 2008, Vol.129 (1-3), p.160-164
• Main Authors: Bor, Doğan, Tukel, Selma, Olgar, Turan, Toklu, Türkay, Aydın, Elif, Akyol, Oya
• Format: Article
• Language: English
• Subjects: Breast - anatomy & histology; Breast - radiation effects; Breast Neoplasms - prevention & control; Female; Humans; Mammography - instrumentation; Mammography - methods; Mammography - statistics & numerical data; Radiation Dosage; Radiographic Image Enhancement - methods; Risk Assessment
• ISSN: 0144-8420; 1742-3406
• DOI: 10.1093/rpd/ncn140

The relationship between the mean glandular dose (MGD) and the compressed breast thickness (CBT) is commonly used for the presentation of mammographic dose survey results and could also be useful for the assessment of individual breast doses retrospectively in case of lack of necessary dosimetric instrumentation. The high data scattering from the best fit reduces the reliability of this technique. The aim of this study was to investigate the accuracy of this relationship using the data collected from a patient survey and phantom experiment. Patients were divided into three different groups according to their breast glandularities, which were predicted from the inspection of previous mammograms. X-ray beam qualities that will be used in patient examinations were determined according to breast thickness and predicted glandularities. The MGD versus CBT relationship for all the examined patients resulted in a poor correlation (R2 = 0.28). This relationship was separately obtained for each glandularity group and also for sub-groups of specific beam qualities. The best correlation (R2 = 0.73) was obtained for the fatty breast group and Mo/Mo combination. A low correlation (R2 = 0.34) was observed in the mid-glandularity group due to inclusion of a wide range of glandularities in this group. In the case of the dense breast group, although the glandularity range was narrow, there were e still high data scattering (R2 = 0.25). This was probably due to the use of Mo/Rh and Mo/Mo combinations. This is validated by obtaining the MGD–CBT relationship specific to Mo/Mo combination (R2 = 0.61).