Energy dependence and angular dependence of an optically stimulated luminescence dosimeter in the mammography energy range

Abstract This study aimed to investigate the energy dependence and the angular dependence of commercially available optically stimulated luminescence (OSL) point dosimeters in the mammography energy range. The energy dependence was evaluated to calculate calibration factors (CFs). The half‐value layer range was 0.31–0.60 mmAl (Mo/Mo 22–28 kV, Mo/Rh 28–32 kV, and W/Rh 30–34 kV at 2‐kV intervals). Mo/Rh 28 kV was the reference condition. Angular dependence was tested by rotating the X‐ray tube from −90° to 90° in 30° increments, and signal counts from angled nanoDots were normalized to the 0° signal counts. Angular dependence was compared with three tube voltage and target/filter combinations (Mo/Mo 26 kV, Mo/Rh 28 kV and W/Rh 32 kV). The CFs of energy dependence were 0.94–1.06. In Mo/Mo 26–28 kV and Mo/Rh 28–32 kV, the range of CF was 0.99–1.01, which was very similar. For angular dependence, the most deteriorated normalized values (Mo/Mo, 0.37; Mo/Rh, 0.43; and W/Rh, 0.58) were observed when the X‐ray tube was rotated at a 90° angle, compared to 0°. The most angular dependences of ± 30°, 60°, and 90° decreased by approximately 4%, 14%, and 63% respectively. The mean deteriorated measurement 30° intervals from 0° to ± 30° was 2%, from ± 30° to ± 60° was 8%, and from ± 60° to ± 90° was 40%. The range of energy dependence in typical mammography energy range was not as much as that in general radiography and computed tomography. For accurate measurement using nanoDot, the tilt needs to be under 30°.


| INTRODUCTION
Mammography is the most effective method to screen for breast cancer. The breast includes skin, glandular tissue, adipose tissue, and areolar tissue. The glandular tissue is highly sensitive to some adverse effects of radiation. 1 A prerequisite in patient dose management is that the benefits of screening should be considerably greater than the risks induced by the use of radiation.
Some point-based dosimeters such as thermoluminescent dosimeters 2 and radio photoluminescent glass dosimeters 3 have been used to directly measure dosage in patient dose management.
The energy range in mammography is very low, compared to that used in general radiography and computed tomography (CT). Therefore, these point-based dosimeters usually require corrections for energy dependence. [4][5][6] Optically stimulated luminescent (OSL) dosimeters were recently introduced as point-based dosimeters into medical and environmental dosimetry. The OSL dosimeter was first adopted for medical dosimetry in radiation therapy. [7][8][9][10] Later, research widened to studying the feasibility of using OSL dosimeters in the diagnostic energy range. [11][12][13][14] The mechanism of optically stimulated luminescence and thermoluminescence are similar processes. 8,15 The structure of the dosimeter is composed of pure crystalline dielectric materials and contains a small quantity of dopants that form crystal-lattice imperfections. These imperfections act as traps for electrons and holes.
Electrons and holes are trapped by these energy traps after exposure to ionizing radiation. When the crystal is stimulated with a lightemitting diode, for example, at a dosimeter readout, the electrons can be ejected out of traps and recombined with holes while emitting characteristic light proportional to the amount of the absorbed radiation dose. 8 Currently, the only material used broadly in OSL dosimeters is aluminum oxide with carbon doping (Al 2 O 3 :C). One type of Al 2 O 3based OSL dosimeter, the nanoDot (Landauer, Inc., Glenwood, IL, USA), is commercially available and is small, robust, and reusable. It has high sensitivity, and its density is near to that of a human body, 11,12,16 making it a realistic choice for point measurements in diagnostic imaging. 13,14 Jursinic et al. 8 reported that OSL dosimeters exhibit high precision and accuracy in measuring a dose, and they have no energy dependence, and no dependence on the irradiation angle in the radiation therapy energy range. Al-Senan and Hatab 11 investigated the feasibility of using commercially available OSL dosimeters in the diagnostic energy range including a part of the mammography energy range. The linearity test showed good linear response with R 2 > 0.99 and the angular dependence showed the maximum variation as a drop of approximately 70% at 90°at only 25 kV in the mammography energy range. The OSL dosimeters also had energy dependence and were recommended to acquire correction factors in the diagnostic energy range, however, the mammography energy range was not investigated. The energy dependence and angular dependence of the OSL dosimeter require further investigation to evaluate the feasibility of their use in the mammography energy range, based on the various tube voltage of each target and filter combinations.
In this study, we present an evaluation of two specific dosimetric characteristics in the mammography energy range of a commercial OSL dosimeter. We investigated energy dependence and angular dependence based on tube voltage of each target and filter combination.

| MATERIALS AND METHODS
The OSL dosimeter system included InLight nanoDot OSL dosimeters (Landauer Inc., Glenwood, IL, USA) and the microStar reader  The CF for energy dependence was defined as 12

2.B | Angular dependence
The variability in the nanoDot response to incident X-ray beams from various angles was investigated. Seven different angles were evaluated with eight nanoDots (one nanoDot was used as a control).
Individual nanoDot sensitivity was corrected to acquire an average signal in counts/mGy. A nanoDot set on hard paper at the rotation isocenter of the X-ray tube (isocenter points, The X-ray tube was rotated by À90°, À60°, À30°, 0°, 30°, 60°, and 90°[ Fig. 3(a)]. The same cycle of irradiation, readout, and bleaching was repeated three times. Signal counts from angled nanoDots were normalized to the 0°signal counts, in which the detector's serial number was facing the beam.

| RESULTS
To confirm the reproducibility of the radiation output and geometric arrangement of the ionization chamber, the ionization chamber was   Table 1 lists the CFs, corresponding

| DISCUSSION
Commercially available OSL dosimeters were tested for the energy dependence and the angular dependence based on the tube voltage of each target and filter combination in the mammography energy range. Energy dependences were observed in approximately 10% of the typical mammography energy range. Angular dependence showed the most deteriorated measurement and suggested that nanoDots were limited to rotate under 30°for the measurement accuracy (under 4%).
The OSL dosimeter almost never showed energy dependence in the therapeutic energy range. 9,16 On the other hand, it has been shown that the OSL dosimeter over-responded to low-energy Xrays. 7,11,12 This can be attributed to the high photoelectric effect of Al 2 O 3 :C at low photon energy values, which raises its mass energy absorption coefficient relative to water. 11 Al-Senan and Hatab 11 reported that the energy dependence in general radiography energy range was between 0.81 and 1.56. The range of energy dependence in this study was lower than that in general radiography energy range. Scarboro et al. 12 reported that the variation in energy dependence for CT in air and a phantom was between + 20% and À15%.
This energy dependence was higher than that of the result in this study. Hsu et al. 5  The angular dependence of the nanoDot was compared between three tube voltage and target/filter combinations. The greatest deterioration was about 60% when the X-ray tube was rotated at a 90°a ngle compared with 0°for all directions. However, the deterioration of angle interval was not constant, the deterioration of over 60°i ncreased more. The lower the effective energy, the higher was the deterioration of the signal counts in the angular dependence. More angled X-rays were filtered out and attenuated before reaching the detector because an angled X-ray travels a longer distance while passing through the plastic case than an X-ray facing the detector's serial number. Al-Senan and Hatab 11 reported that the angular dependence of nanoDot dosimeters showed variations as high as 70% in mammography and were close to the result of this study, and angular dependence was the highest for other modalities, including therapeutic settings (under 5%), 17 general radiography (under 40%), 11 and CT (under 11%). 12 In a breast tomosynthesis system, effective energy is higher than those of mammography and the Xray tube is rotated from 15°to 50°. 20

| CONCLUSIONS
The commercial OSL dosimeter evaluated the 2 specific dosimetric characteristics, which were energy and angular dependence, based on the tube voltage of each target and filter combinations in the mammography energy range. Energy dependence in the typical mammography energy range of the OSL dosimeter was lower than that in general radiography and CT, and was equal to that of thin film-thermoluminescent dosimeters. Angular dependence showed the least measurement accuracy for all target and filter combinations. In energy dependence and angular dependence, the OSL dosimetercalibrated typical mammography energy was suggested to be used for accurate measurement at under 30°of the tilt.