Comparative study of computational dosimetry involving homogeneous phantoms and a voxel phantom in mammography: a discussion on applications in constancy tests and calculation of glandular dose in patients*

OBJECTIVE: To compare data regarding dosimetry and photons fluence in different breast phantoms, discussing constancy tests and dosimetry applied to mammography. MATERIALS AND METHODS: Different homogeneous breast phantoms and one anthropomorphic voxel phantom were developed for collection of data regarding total absorbed dose in the phantom, absorbed dose in the glandular tissue material-equivalent, absorbed dose and photons fluence at different depths in the phantoms. A simulated ionization chamber collected the entrance skin kerma. Target-filter combinations (Mo-30Mo and Mo-25Rh) were studied for different accelerating potentials of 26 kVp to 34 kVp. RESULTS: As compared with the voxel phantom, the normalized glandular dose resulted in differences from –15% to –21% for RMI, –10% for PhantomMama, and 10% for the Barts and Keithley models. The half-value layer variation was generally < 10% for all the sensitive volumes. CONCLUSION: The phantom proposed by Dance is recommended for evaluating the glandular dose and normalized glandular dose in a standard breast. Homogeneous phantoms should be utilized for constancy tests in dosimetry, but they are not appropriate for estimating dosimetry in actual patients.