Keywords:
Breast, Mammography, Dosimetry, Physics
Authors:
F. Cavagnetto1, R. Bampi2, M. Calabrese1, F. Chiesa1, G. Taccini1; 1Genova/IT, 2Torino/IT
DOI:
10.1594/ecr2011/C-1963
Methods and Materials
Fig.
BestMedical TN-1002RDM microMOSFET,
high sensitivity bias power supply,
were calibrated for the digital mammography system Hologic Selenia Dimensions for all clinical parameters [3]: target-filter combinations W/Rh (28-31 kVp) and W/Ag (30kVp) for FFDM and W/Al (30-35 kVp) for DBT.
Linear response with mAs has been tested varying kVp and anode-filter.
To avoid MOSFET perturbation of the image,
the dosimeter is placed at a fix chest wall-to-nipple distance outside compressed breast: to determine breast entrance skin air kerma (BESAK) a correction factor is used.
MOSFET sensitivity mV/cGy ranging between 24 and 30,
with error 6%; R2 (linear correlation coefficient) response vs mAs is >0,98.
Dosimeters are fixed under the compressor plate before mammography,
the reading is unique at the end of the exam,
so in vivo dosimetry don’t perturb or delay mammography.
Using acquisition parameters and calibration data,
an Excel worksheet provide BESAK for each view both FFDM or DBT and COMBO modality.
The methodology has been tested with a PMMA phantom and with 50 patients,
the differences between measured and calculated values are <7% in FFDM and DBT,
<8% in COMBO modality,
for compressed breasts thicker than 30 mm,
otherwise the dosimeter is instable and the error can be higher than 15%.