Characterization of a fiber-optic-coupled radioluminescent detector for application in the mammography energy range.

Fiber-optic-coupled radioluminescent (FOC) dosimeters are members of a new family of dosimeters that are finding increased clinical applications. This study provides the first characterization of a Cu doped quartz FOC dosimeter at diagnostic energies, specifically across the range of x-ray energies and intensities used in mammographies. We characterize the calibration factors, linearity, angular dependence, and reproducibility of the FOC dosimeters. The sensitive element of each dosimeter was coupled to a photon counting photomultiplier module via 1 m long optical fibers. A computer controlled interface permitted real-time monitoring of the dosimeter output and rapid data acquisition. The axial-angular responses for all dosimeter models show nearly uniform response without any marked decrease in sensitivity. However, the normal-to-axial angular response showed a marked decrease in sensitivity of about 0 degrees C and 180 degrees C. In most clinical applications, appropriate dosimeter positioning can minimize the contributions of the varying normal-to-axial response. The FOC dosimeters having the greatest sensitive length provided the greatest sensitivity, with greatest to lowest sensitivity observed for 4.0, 1.9, 1.6, and 1.1 mm length sensitive elements. The average sensitivity of the dosimeters varies linearly with sensitive volume (R2=95%) and as a function of tube potential and target/filter combinations, generally exhibiting an increased sensitivity for higher energies. The dosimeter sensitivity as a function of tube potential had an average increase of 4.72 +/- 2.04% for dosimeter models and three target-filter combinations tested (Mo/Mo, Mo/Rh, and Rh/Rh) over a range of 25-31 kVp. All dosimeter models exhibited a linear response (R2 > or = 0.997) to exposure for all target-filter combinations, tube potentials, and tube current-time product stations evaluated and demonstrated reproducibility within 2%. All of the dosimeters examined in this study provided a response adequate for the accurate measurement of doses in clinical mammography applications.