- Email: [email protected]
Characterization of OSL dosimeters for use in dose assessment in Computed Tomography procedures
Physica Medica 47 (2018) 16–22
Contents lists available at ScienceDirect
Physica Medica journal homepage: www.elsevier.com/locate/ejmp
Original paper
Characterization of OSL dosimeters for use in dose assessment in Computed Tomography procedures Louise Giansante, Josilene C. Santos, Nancy K. Umisedo, Ricardo A. Terini, Paulo R. Costa
T
⁎
Nuclear Physics Department, Institute of Physics, University of São Paulo, São Paulo, Brazil
A R T I C L E I N F O
A B S T R A C T
Keywords: OSL Dosimetry Instrumentation Computed Tomography Radiology
This study describes the characterization of an Al2O3:C OSLD (Landauer’s Luxel™ tape) for dose evaluation in Computed Tomography. The irradiations were conducted using both a constant potential X-ray equipment and a 64-slice clinical CT scanner, and the readouts were performed using a Risø TL/OSL reader. The following aspects were studied: batch homogeneity, energy response, linearity of dose response, reproducibility, reusability, and effect of uncertainties with the normalization of OSL signals per their response to beta radiation. A group of 330 dosimeters from the 452 irradiated with the same dose presented OSL signals within the interval of 4.7% from the average. The dosimeters presented energy-dependent response in good agreement with results found in the literature. The air kerma response of the OSL signal showed a linear trend for both the constant potential X-ray device and the clinical CT scanner, with differences in their slopes of approximately 10%. Reproducibility, reusability, and effect of beta normalization were analyzed by separating 72 dosimeters in 3 groups. The results obtained in this study together with those of previous works indicate that this type of dosimeter is adequate for dose evaluation in CT clinical applications.
1. Introduction Since the development of the first Computed Tomography (CT) equipment in 1970, this diagnostic imaging modality has been rapidly expanding [1]. The radiation dose absorbed by patients due to this technique has become a concern among radiologists, researchers, and manufacturers [2–4], leading to the development of different methods to evaluate the absorbed dose in such examinations [5]. Ionization chambers (IC), thermoluminescence (TL) and, more recently, optically stimulated luminescence (OSL) dosimetry, for instance, have been widely used in order to estimate doses in vivo, in post-mortem subjects, and in phantoms. Lavoie et al. [6], for instance, proposed an experimental method to assess the nanoDot Al2O3:C dosimeter (Landauer, Inc.), aiming to apply this system in CT procedures. In the study conducted by Funama et al. [7], the authors measured dose profiles in radiosensitive organs in thorax and breast regions due to CT Coronary Angiography (CTCA) procedures. Therapeutic and diagnostic beams were assessed in the low-medium energy range by Spasic & Adam [8]. More recently, the optimal bleaching conditions were assessed with fully filled traps Al2O3:C dosimeters in order to evaluate changes in their dose sensitivity [9]. In the review proposed by Alaei & Spezi [10], the authors present
⁎
several methods used to calculate dose due to CBCT procedures, including a variety of phantoms and dosimeters. The nanoDot Al2O3:C from Landauer was also applied to measure out-of-field doses in radiotherapy, and results showed good agreement with results obtained with other types of dosimeters [11]. Takegami and colleagues [12] have used a modern 320-slice CT scanner to evaluate the entrance surface dose (ESD) in phantoms and in patients. Results reported by these authors and others indicate that this dosimetry system offers many advantages compared with other systems, such as efficiency, accuracy, linearity, and good spatial resolution, besides keeping the signal after the readouts. The present study proposes an experimental methodology for the characterization in the diagnostic range of an Al2O3:C optically stimulated dosimeter (Landauer, Inc.). The objective of the present study was to verify the applicability of this type of dosimeter in CT organ dose assessment using adult and pediatric anthropomorphic phantoms in clinical CT machines [13]. The authors demonstrate the potentiality of this technology for this kind of purpose, but emphasize that although several of the OSL properties are well documented in the literature, in special for personal dosimetry, its applicability for other purposes were not still exhaustively validated by scientific methodologies. This paper intends to contribute for this validation process.
Corresponding author at: 1371 Matão St., 05508-090 São Paulo, SP, Brazil. E-mail addresses: [email protected] (L. Giansante), [email protected] (P.R. Costa).
https://doi.org/10.1016/j.ejmp.2018.02.009 Received 4 October 2017; Received in revised form 10 February 2018; Accepted 13 February 2018 1120-1797/ © 2018 Published by Elsevier Ltd on behalf of Associazione Italiana di Fisica Medica.
Physica Medica 47 (2018) 16–22
L. Giansante et al.
2. Materials and methods
the literature [23,24]. For TL measurements each dosimeter was heated from the room temperature until 350 °C at a constant rate of 10 °C per second [15]. The TL signal was extracted by integrating the TL curve until 350 °C.
The following characteristics of OSLDs were investigated: batch homogeneity, energy response, linearity of dose response, reproducibility, reusability, and beta response normalization. The Landauer Luxel™ tape (Landauer, Inc., Glenwood, USA), from a 73 m roll measuring 20.0 mm in width and 0.3 mm in thickness, produced with carbon-doped aluminum oxide crystals converted into powder [14], was used in the present study. The tape was fractionated by the authors into disks measuring 3 or 5 mm in diameter, according to their application in the present study. A group of pre-selected thermoluminescent dosimeters (TLDs), with 3 mm × 3 mm × 1 mm, composed of LiF:Mg,Ti (TLD-100, Harshaw Chemical Company, OH, USA), was simultaneously irradiated along with the 3 mm OSLDs in a set of control measurements for comparison of results. Before each exposure, the OSLDs were optically treated (bleaching) for 8 h under fluorescent light (Digilight, ATEK, São Paulo, Brazil). Similarly, the TLDs were heated (annealing) at 400 °C for 1 h followed by 100 °C for 2 h. Both procedures aim to erase the effects of previous irradiations, stabilizing the sensitivity and the background of the dosimeters so that their properties remain invariable throughout usage [15]. Two X-ray sources were used in the present study: a high precision X-ray machine used for personal dosimeter calibration laboratory was adopted for the basic OSLD characterization tests and a clinical machine was used for linearity tests. The irradiations for characterization purposes were performed using a constant potential X-ray tube MCN 421 (Philips, Germany). RQR and RQT X-ray beam standards [16,17] were validated and characterized in this equipment [18] and used during the procedures described in the present study. A clinical 64-slice CT scanner, Brilliance 64 (Philips, Germany), from the Institute of Radiology of the School of Medicine of the University of São Paulo (InRad/FMUSP), was used in an additional set of linearity tests in order to verify the degree of similarity between the two machines. Two ionization chambers (ICs) were used. A 30 cc IC, model 233610576 (PTW, Freiburg, Germany), which is used as a reference chamber for absolute dosimetry in calibration laboratories [19], was used in the batch homogeneity and energy response tests. This chamber was coupled to an electrometer Unidos E (PTW, Freiburg, Germany) and it was chosen because of its high accuracy and signal stability. The second chamber was a 0.6 cc IC (Radcal Corporation, Monrovia, CA, USA), model 10X5-0.6. This device is currently applied for dose measurements in multi-slice Computed Tomography scanners [20,21] and it was chosen because its small volume allowed high accurate measurements in the isocenter of the clinical CT X-ray beam. It was used in the linearity, reproducibility, reusability and beta normalization effect tests. This chamber was coupled to an electrometer model 9010 (Radcal Corporation, Monrovia, CA, USA). Both ICs and electrometers were calibrated by an SSDL (Secondary Standard Dosimetry Laboratories) [16,17]. The experimental arrangement, combining X-ray source and IC, is described in the following paragraphs according to each experiment. A Risø TL/OSL reader, model DA-20 (DTU Nutech. Inc., Roskilde, Denmark), with a built-in 90Sr/90Y beta source and blue light stimulation, was used to read the information from the dosimeters. A bi-alkali photomultiplier tube (model 9235QB) was used to detect both OSL and TL signals. An ultraviolet transmitting broad-band pass filter (Hoya U340) and a blue filter pack (Schott BG-39 + BG-3) were used in front of the photomultiplier tube for OSL and TL measurements respectively [22]. Continuous blue light stimulation was applied to each disk for 90 s, and the OSL signal was extracted from the initial signal of each OSL curve. After performing the first readout, a group of OSLDs were irradiated with beta radiation for 2 s and read again. Their responses to beta radiation were used to individually normalize the OSL signals from the X-ray radiation, decreasing the uncertainties, as once suggested in
2.1. Batch homogeneity A group of 452 dosimeters was irradiated using an MCN 421 X-ray tube to evaluate the batch homogeneity of the OSLD response to the same beam quality. The RQT 9 X-ray beam quality was used, as it usually corresponds to the reference radiation quality for calibration of instruments utilized in Computed Tomography dosimetry. The 30 cc IC was positioned 5 m away from the X-ray tube, and the air kerma measured was 20.4 ± 0.3 mGy. 2.2. Energy response OSL energy response was evaluated for eight different standard Xray beam qualities: RQR 2, RQR 4, RQR 5, RQR 6, RQR 8, RQT 8, RQT 9, and RQT 10. The RQR series of radiation qualities represent the beam incident on the patient in general radiography, whereas the RQT series simulate the unattenuated beam used in CT procedures [16]. Differences between each series are due to the X-ray tube voltage and filtration and, therefore, the effective energy of the X-ray beam. Eight groups of three OSLDs were irradiated using the MCN 421 X-ray tube, and the corresponding air kerma values were measured with the 30 cc IC positioned 5 m away from the focal spot of the X-ray tube. The OSL signal of each quality, after readout, was evaluated by calculating the mean of three dosimeter responses and the uncertainties were obtained as the standard deviation of the mean (k = 1).1 The effective energy, Eeff, of each X-ray beam quality was then determined (Eq. (1)) and their values were related to the OSL signal obtained for each quality.
⎛ μ ⎞ (Eeff ) = ln(2) HVL×ρAl ⎝ ρ ⎠Al ⎜
⎟
(1)
The HVL for each X-ray beam quality was determined experimentally, based on data provided by the IAEA Technical Series Report 457 [16] and the aluminum density was determined with data provided by the National Institute of Standards and Technology (NIST) [25]. Therefore, with ρAl = 2.699 g/cm3 and values of HVL, mass attenuation
() μ
coefficient ⎛ ρ ⎞ for each X-ray beam quality were determined. The Al ⎠ ⎝ XCOM program [26] was used to find the energies associated to each
()
μ ρ Al
values, adopted as effective energy corresponding to each HVL
value. Table 1 summarizes tube voltage (kV), external filtering, HVL, and Eeff for each X-ray beam quality.
2.3. Linearity of response The linearity of the OSL responses to the incident air kerma were first assessed using the Philips 64-slice CT scanner and the Radcal 0.6 cc IC. The IC and the dosimeters were positioned in the center of the gantry on an acrylic holder attached to a support (Fig. 1). Five groups of three unused OSLDs were irradiated along with five groups of three TLDs. These experiments used a tube voltage of 120 kV and five different values of tube current-time product ranging from 30 mAs to 400 mAs, using different number of rotations. The resulting IC readings for these settings ranged from 3.9 mGy to 51.8 mGy. After these experiments, the OSLDs were read and irradiated with the beta source of the Risø reader, and their responses to beta radiation were used to individually normalize their responses to the X-ray. 1 It is reasonable to assume that the values obtained with the three dosimeters follow a Normal distribution, thus the proper t value was applied to the standard deviation of the mean for k = 1 (68.3% of confidence).
17
Physica Medica 47 (2018) 16–22
L. Giansante et al.
Table 1 X-ray beam qualities RQR and RQT series used to assess the energy response of the analyzed dosimeters. X-ray tube voltage (kV), external copper filtration (mm Cu), nominal first HVL (mm Al), and determined effective energy are presented for each radiation quality. Radiation quality
X-ray tube voltage1 [kV]
External filtration1 [mm Cu]
Nominal first HVL [mm Al]
Determined effective energy [keV]
RQR 2 RQR 4 RQR 5 RQR 6 RQR 8 RQT 8 RQT 92 RQT 10
40 60 70 80 100 100 120 150
– – – – – 0.20 0.25 0.30
1.42 2.19 2.58 3.01 3.97 6.90 8.40 10.10
25.3 29.7 31.3 33.2 37.9 49.7 56.1 64.2
Table 3 OSL groups considered for the reproducibility and reusability tests. Group
Cycle
Group 1 Group 2 Group 3
Bleaching, reading, irradiation, reading Bleaching, irradiation, reading Bleaching, irradiation, reading, irradiation with beta source, reading
generate calibration curves, which can be used, for example, to evaluate doses due to CT procedures. The techniques applied in each situation are summarized in Table 2.
2.4. Reproducibility, reusability, and beta normalization effect Reproducibility and reusability of OSLDs were evaluated using the 0.6 cc IC positioned 1 m away from the focal spot of the MCN 421 X-ray tube and X-ray beam quality RQT 9. Dosimeters were fractionated into
1
Data retrieved from the IAEA Technical Series Report 457 [16]. Reference radiation quality usually applied in calibration of equipment for CT dosimetry; therefore, data were normalized for this quality. 2
Fig. 1. Left: The ionization chamber at the isocenter of the gantry of the 64-slice Philips Brilliance 64 CT Scanner (Philips, Germany). Right: The 0.6 cc ionization chamber placed on an acrylic holder. Air kerma was measured at this position and then each group of TLD and OSLD was positioned at the same place.
disks measuring 5 mm in diameter for these experiments. A tube voltage of 120 kV, tube current of 9 mA and exposure time of 50 s were used, so that the ionization chamber reading was 20.3 ± 0.3 mGy. Four exposures were performed in a one-week period using the same parameters, with 72 dosimeters separated in three distinct groups of 24 each. Groups were separated according to Table 3, in order to choose the most accurate method. The same cycle was performed for each group four times. Reproducibility test aimed to evaluate the precision of dosimeter responses under the same irradiation conditions. Reusability test aimed to verify the efficacy of the bleaching and the possibility of reusing the dosimeters in subsequent irradiations.
Table 2 Parameters used to assess linearity of OSLD responses in distinct situations. At the laboratory, dosimeters and IC were placed 1 m away from the X-ray tube focal spot. At the hospital, dosimeters and IC were placed in the center of the gantry of the Philips Brilliance 64 CT Scanner. Source/ chamber
Tube voltage (kV)
Tube current-time product (mAs)
Air Kerma (mGy)
TLD group
OSLD group
CT Scanner/ 0.6 cc
120
30 100 150 200 400
3.92 ± 0.09 13.05 ± 0.28 19.55 ± 0.41 26.01 ± 0.56 51.83 ± 1.10
1 2 3 4 5
1 2 3 4 5
MCN 421 Xray tube/ 0.6 cc
120
90 270 540 1080 2700
4.02 ± 0.10 12.90 ± 0.31 26.25 ± 0.63 53.0 ± 1.3 133 ± 3
1 2 3 4 5
1 2 3 4 5
3. Results and discussion 3.1. Batch homogeneity The OSL responses to the X-ray irradiation presented an approximate Gaussian behavior (Fig. 2), which was verified by AndersonDarling statistics performed with the Wolfram Mathematica® 10.0 software. A group of 330 dosimeters from the 452 irradiated disks (73% of the initial batch) presented OSL signals within ± 4.7% around the average (85021 ± 4018 counts). This demonstrates that most of the dosimeters respond in a similar way under the same irradiation conditions. Given the possible differences in dosimeters shape, size, and uniformity due to their small size and the manually fractioning process, a 4.7% variation was considered small. The uniformity of the OSL Luxel™
In another set of measurements, five groups of three OSLDs and five groups of three TLDs along with the 0.6 cc IC were positioned one meter away from the focal spot of the MCN 421 X-ray tube. These experiments used a tube voltage of 120 kV, tube current of 9 mA, and five different values of exposure time ranging from 10 s to 300 s. The resulting IC readings for these settings ranged from 4 mGy to 133 mGy. These measurements aimed to verify the applicability of the CT standard qualities with the MCN 421 X-ray tube for irradiating OSLD or TLD to
18
Physica Medica 47 (2018) 16–22
L. Giansante et al.
author indicated that the aluminum oxide powder in the OSL strip was not homogeneously distributed. However, this non-uniformity did not affect the results of CT dose profiles measurements [27]. Recently, Han et al. [28] evaluated the characteristic of a 5-mm Landauer nanoDot for an 80 kV cephalometric exposure. Homogeneity was evaluated for a batch of 25 dosimeters, and the authors reported a variation of 1.67%.
3.2. Energy response The responses of the OSLDs followed a non-uniform energy dependence (Fig. 3), similar to those found in the literature [6,29–31]. Gasparian et al. [30] investigated the responses of two types of OSLDs: Al2O3:C single crystals of 5 mm in diameter and 0.9 mm in thickness and Al2O3:C Luxel dosimeters of 7 mm in diameter and 0.3 mm in thickness. Malthez et al. [31] conducted similar analysis with five different TL/OSL materials, including the Luxel OSLD cut in 7 mm of diameter and 0.3 mm in thickness. To compare the results of this study with those previously published, the authors provided their data, which was normalized to the results of this paper. The results verified by those authors, after normalization, and the ones obtained in this work presented similar energy-dependent response (Fig. 3). In CT procedures, 120 kV is the most commonly applied voltage for adult protocols [6]; however, lower tube voltages such as 80 kV, 90 kV, or 100 kV are also routinely used, depending on the clinical procedure and on the anatomy of the patient. For each case, there is a different value of HVL, which makes the evaluation of the behavior of the dosimeters necessary under these circumstances [6]. Such energy dependence in diagnostic energy range can be corrected with proper calibration.
Fig. 2. Distribution of the OSLD responses. 452 unused dosimeters were irradiated at the same time using the same technique.
3.3. Linearity of response The OSL signal dependence to the air kerma is presented in Fig. 4. An expected linear dependence was found when the 0.6 cc IC was irradiated by the MCN 421 X-ray tube and the Philips Brilliance 64 CT scanner together with TLDs and OSLDs. However, the slope of the linearity curve for the X-ray tube was 10% higher than the one obtained with the CT scanner for the TLDs and 8% higher for the OSLDs (Fig. 4). Those differences might be related, for instance, to the degree of similarity of the X-ray beam qualities of these two equipment, flux dependence, and to scattering conditions. The results of the linearity evaluation are summarized in Table 4.
Fig. 3. OSLD energy responses obtained in this study for eight different X-ray beam qualities. The mean OSL signal obtained for each quality was normalized to the one verified for the RQT 9 X-ray beam quality. The other two symbols correspond to the energy responses for the Al2O3:C Luxel dosimeters obtained in distinct works. These data were normalized to the data found in this work for comparison.
tape was investigated by Gasparian [27]. The author performed this analysis in OSL strips cut from the original tape irradiated using a linear accelerator with photon energy of 6 MeV. Results reported by this
Fig. 4. Comparison of the linearity responses of the OSLDs (left) and TLDs (right) to the air kerma measured by a 0.6 cc IC in the MCN 421 X-ray tube and in the CT scanner. The OSL signal was normalized by their responses to beta radiation, as previously explained (Section 2.3).
19
Physica Medica 47 (2018) 16–22
L. Giansante et al.
Table 4 OSLD and TLD linearity response parameters obtained for distinct situations for 120 kV. Source
Ionization chamber (IC)
Dosimeter
Distance source-IC/dosimeter [m]
Slope (counts/mGy)
R2
X-ray tube CT scanner X-ray tube CT scanner
0.6 cc 0.6 cc 0.6 cc 0.6 cc
OSLD OSLD TLD TLD
1 Center of the gantry 1 Center of the gantry
0.141 ± 0.0021 0.135 ± 0.0021 (73.5 ± 1.8) × 103 (66.2 ± 0.9)8103
0.9986 0.9990 0.9992 0.9975
1
OSL signal normalized to beta radiation.
Fig. 5. Dosimeter responses per irradiation for groups 1 and 2 and normalized OSL signal for group 3. For each data set, whiskers present the variation range (minimum and maximum values), the box contains 50% of data (75%–25%), the horizontal line inside the box represents the median value of the responses and the square represents the mean value of the data. Groups 1 and 2 presented outliers.
respectively, 2.4% and 0.26%. The comparative evaluation between the results obtained with dosimeters from Groups 1, 2, and 3 highlights the effect of normalizing each OSL signal (SX-ray) by their signal obtained after the irradiation with a beta source (Sβ). With the “normalized OSL signal”, dependence with the homogeneity of the crystals in the dosimeter, caused by the manual process of fractionating the tape, or stability of the reader, for instance, are taken into account and individually corrected. In Fig. 5, dosimeter responses from three groups are presented in box plots per irradiation. The absence of outliers in the responses from Group 3 highlights the efficacy of individually normalizing the OSL signals per their responses to beta radiation. Dosimeter responses from Groups 1 and 2 presented outliers in all readings.
3.4. Reproducibility, reusability, and beta normalization effect After irradiating the three groups under the same conditions, mean, standard deviation and coefficient of variation (CV) for each dosimeter group were recorded. OSL signals from each group were normalized so that their CVs could be comparable. This study indicated a variation among irradiations from 3% to 5% for Group 1, from 4% to 5% for Group 2 and from 1% to 2% for Group 3. Similarly, during the process of irradiating and erasing the 24 dosimeters from the three groups, each dosimeter was identified and their signal was individually recorded along the readings to assess the reusability. After performing 4 cycles (Table 3), CV values varied from 1% to 4% for dosimeters from Groups 1 and 2 and from 1% to 2% for dosimeters from Group 3. In a similar analysis, Musa and colleagues [32] measured the CV of ten repeated measurements using two nanoDots. In that study, the authors used a higher dose (10 and 1 Gy), and the CV values encountered were, 20
Physica Medica 47 (2018) 16–22
L. Giansante et al.
and Development) for the financial support (process 132554/2015-1), and the Institute of Radiology of the School of Medicine of the University of Sao Paulo (InRad/FMUSP) for the partnership (CAAE 55420616.3.0000.0068). We thank Malthez and colleagues, and Gasparian and colleagues for providing their data that we used for comparison. We also thank the CNPq/FAPESP for funding the project by INCT – Metrology of Ionizing Radiation in Medicine (process 2008/ 57863-2).
4. Conclusions This study presented a methodology for the characterization of the Al2O3:C Luxel OSLD tape manufactured by Landauer, Inc., manually fractioned for CT applications. The results of some tests were compared with those obtained with the LiF:Mg,Ti thermoluminescent dosimeters and with the results found in the literature regarding works with similar dosimeters. The dosimeters responses were analyzed for homogeneity, linearity with the incident air kerma, reproducibility, reusability, and energy-dependence to distinct effective energies. The batch homogeneity presented a Gaussian behavior with 4.7% of standard deviation, which demonstrates that most of the dosimeters in the same batch respond in a similar way. However, the normalization of the OSL signals to their responses to beta radiation proved to be efficient in decreasing uncertainties related to differences among dosimeters and by non-stability of the reader. The coefficient of variation of the responses decreased by at least 3%, thus proving the efficiency of such analysis. Reproducibility and reusability tests conducted in this study demonstrated that this kind of dosimeter can be reused in subsequent radiations for low-dose procedures. There were no relevant differences in dosimeter responses after 4 consecutive irradiations and the 8-h bleaching process proved to be efficient. When working with higher doses, however, a more extensive characterization might be needed to ensure doses are properly erased. Moreover, individual readings of each dosimeter can present a higher variation, thus it is recommended to use more than one dosimeter per exposure, and to include the standard deviation of their mean as a component of the uncertainty for a more reliable result. Linearity was assessed up to 140 mGy with a 0.6 cc IC, a constant potential X-ray tube, and a 64-slice CT scanner. This is sufficient for CT procedures, but a more extensive range of characterization is needed for other clinical applications with higher doses. Moreover, the linear response indicates that these curves can be applied as calibration curves to convert dosimeter readings into air kerma values. These experiments aimed to evaluate the applicability of the CT standard qualities with the MCN 421 X-ray tube to generate calibration curves. The RQT standard qualities are obtained to reach a specific nominal HVL by adding an external filtration to simulate the non-attenuated beam used in CT procedures [16]. If the X-ray tube and the CT scanner beam qualities were the same, one could not expect discrepancies observed in Fig. 4. Thus, from the obtained results, one can infer that it is preferable to calibrate dosimeters with the same equipment used in the measurements. The energy-dependent response of this kind of dosimeter can be explained in terms of the effective atomic number (Zeff) of Al2O3:C, which is 11.28 [33]. It means that the total photoelectric effect crosssection of this material is higher than that of materials such as soft tissue, air, or water, especially for the low-energy photons typically found in radiology [31]. For such radiation, the energy deposited in the dosimeter is higher than the energy deposited in biological tissues for the same X-ray fluence [31,33]. As a consequence, these dosimeters present overresponse to low-energy x-rays, typically found in radiology [34,35]. Finally, some advantages of these dosimeters over the TLD can be highlighted. The OSLD optical readout process is simpler than the TLD heating readout process, because it eliminates the need of a reliable and reproducible heating system [36]. Furthermore, considering that Landauer provided this material as a tape, the user can fractionate it according to the desired application. In a subsequent work this type of dosimeter was applied in anthropomorphic phantoms to evaluate organ doses due to CT examinations as an alternative to the TLDs [13].
References [1] Rehani MM, Bongartz G, Kalender W, Golding SJ, Gordon L, Murakami T, et al. Managing patient dose in computed tomography. Ann ICRP 2000;30:7–45. [2] Brenner DJ, Hall EJ. Computed tomography – an increasing source of radiation exposure. N Engl J Med 2007;357:2277–84. http://dx.doi.org/10.1056/ NEJMra072149. [3] IAEA. Dosimetry in diagnostic radiology for paediatric patients. Vienna: International Atomic Energy Agency (IAEA Human Health Series No. 24); 2014. [4] Paul J, Chacko A, Saccomandi P, Vogl TJ, Nour-Eldin N-EA. Effect of diagnostic cone-beam computed tomography protocols on image quality, patient dose, and lesion detection. Phys Med 2016;32:1575–83. http://dx.doi.org/10.1016/j.ejmp. 2016.11.111. [5] Longo M, Genovese E, Orlandi C, Donatiello S, Cassano B, Colajacomo M, et al. A novel method for CT dosimetry with a suspended phantom setup. Phys Med 2017;40:122–9. http://dx.doi.org/10.1016/j.ejmp.2017.07.006. [6] Lavoie L, Ghita M, Brateman L, Arreola M. Characterization of a commerciallyavailable, optically-stimulated luminescent dosimetry system for use in computed tomography. Health Phys 2011;101:299–310. http://dx.doi.org/10.1097/HP. 0b013e31820f8e0e. [7] Funama Y, Taguchi K, Utsunomiya D, Oda S, Murasaki H, Yamashita Y, et al. Dose profiles for lung and breast regions at prospective and retrospective CT coronary angiography using optically stimulated luminescence dosimeters on a 64-detector CT scanner. Phys Med 2012;28:76–82. [8] Spasic E, Adam JF. Optically stimulated luminescence for diagnostic and therapeutic low to medium energy X-ray beams experimental dosimetry. Phys Med 2013;29:e42. [9] Park S-Y, Choi CH, Park JM, Chun M, Han JH, Kim J. Sensitivity and stability of optically stimulated luminescence dosimeters with filled deep electron/hole traps under pre-irradiation and bleaching conditions. Phys Med 2017;38:81–7. [10] Alaei P, Spezi E. Imaging dose from cone beam computed tomography in radiation therapy. Phys Med 2015;31:647–58. http://dx.doi.org/10.1016/J.EJMP.2015.06. 003. [11] Harrison R. Out-of-field doses in radiotherapy: input to epidemiological studies and dose-risk models. Phys Med 2017;42:239–46. [12] Takegami K, Hayashi H, Yamada K, Mihara Y, Kimoto N, Kanazawa Y, et al. Entrance surface dose measurements using a small OSL dosimeter with a computed tomography scanner having 320 rows of detectors. Radiol Phys Technol 2017;10:49–59. http://dx.doi.org/10.1007/s12194-016-0366-1. [13] Giansante L. Organ doses evaluation for computed tomography procedures using TL and OSL dosimeters. University of São Paulo; 2017. [14] Yukihara EG, McKeever SWS. Optically stimulated luminescence (OSL) dosimetry in medicine. Phys Med Biol 2008;53:R351–79. http://dx.doi.org/10.1088/00319155/53/20/R01. [15] Furetta C. Questions and answers on thermoluminescence (TL) and optically stimulated luminescence (OSL). World Scientific; 2008. [16] IAEA. Dosimetry in diagnostic radiology: an international code of practice. Vienna: International Atomic Energy Agency (Technical Report Series No.457); 2007. [17] IEC. Medical diagnostic X-ray equipment – radiation conditions for use in the determination of characteristics (1994 Standard IECS No. 61267). Int Electrotech Comm 2005. [18] Burger A, Costa PR. Implementação e Validação de Qualidades de Feixes de Raios X Segundo o Código de Prática da IAEA. 2012. [19] PTW. Cylinder stem ionization chamber 23361. http://www.ptw.de/cylinder_stem_ chamber.html; 2017. 52. [20] AAPM. Comprehensive methodology for the evaluation of radiation dose in X-ray computed tomography AAPM task group 111 (AAPM report 111). Coll Park, MD: Am Assoc Phys Med; 2010. [21] Radcal. 10 × 5 series ion chambers – technical specifications 2013; 2017. [22] Nutech DTU. Guide to the Risø TL/OSL reader. DTU Nutech; 2015. [23] Yukihara EG, Yoshimura EM, Lindstrom TD, Ahmad S, Taylor KK, Mardirossian G. High-precision dosimetry for radiotherapy using the optically stimulated luminescence technique and thin Al2O3:C dosimeters. Phys Med Biol 2005;50:5619–28. http://dx.doi.org/10.1088/0031-9155/50/23/014. [24] Yukihara EG, McKeever SWS. Optically stimulated luminescence: fundamentals and applications. John Wiley & Sons; 2011. [25] NIST. XCOM: photon cross sections database 1998. 2017. [26] Berger MJ, Hubbell JH, Seltzer SM, Chang J, Coursey JS, Sukumar R, et al. NIST standard reference database 8 (XGAM). XCOM phot cross sect database; 2008. [27] Gasparian PBR. Methodological developments for application of optically stimulated luminescence (OSL) in medical dosimetry. Oklahoma State University; 2009. [28] Han SC, Kim KB, Jung H, Ji Y, Park S. Assessment of diagnostic multileaf collimator for cephalometric exposure reduction using optically stimulated luminescent
Acknowledgments The authors thank CNPq (Brazilian National Council for Research 21
Physica Medica 47 (2018) 16–22
L. Giansante et al.
[29] [30]
[31]
[32]
dosemeters. Radiat Prot Dosimetry 2016;174:102–8. http://dx.doi.org/10.1093/ rpd/ncw080. Al-Senan RM, Hatab MR. Characteristics of an OSLD in the diagnostic energy range. Med Phys 2011;38:4396–405. http://dx.doi.org/10.1118/1.3602456. Gasparian PBR, Vanhavere F, Yukihara EG. Evaluating the influence of experimental conditions on the photon energy response of Al2O3:C optically stimulated luminescence detectors. Radiat Meas 2012;47:243–9. http://dx.doi.org/10.1016/j. radmeas.2012.01.012. Malthez ALMC, Freitas MB, Yoshimura EM, Button VLSN. Experimental photon energy response of different dosimetric materials for a dual detector system combining thermoluminescence and optically stimulated luminescence. Radiat Meas 2014;71:133–8. http://dx.doi.org/10.1016/j.radmeas.2014.07.018. Musa Y, Hashim S, Karim MKA, Bakar KA, Ang WC, Salehhon N. Response of optically stimulated luminescence dosimeters subjected to X-rays in diagnostic energy
[33]
[34]
[35]
[36]
22
range. J Phys Conf Ser 2017;851:12001. http://dx.doi.org/10.1088/1742-6596/ 851/1/012001. Bos AJJ. High sensitivity thermoluminescence dosimetry. Nucl Instrum Methods Phys Res Sect B 2001;184:3–28. http://dx.doi.org/10.1016/S0168-583X(01) 00717-0. Mobit P, Agyingi E, Sandison G. Comparison of the energy-response factor of LiF and Al2O3 in radiotherapy beams. Radiat Prot Dosimetry 2006;119:497–9. http:// dx.doi.org/10.1093/rpd/nci676. Perks CA, Yahnke C, Million M. Medical dosimetry using optically stimulated luminescence dots and microStar readers. Proceedings of IRPA12: 12. Congress of the International Radiation Protection Association: Strengthening Radiation Protection Worldwide - Highlights. Global Perspective and Future Trends; 2010. Bøtter-Jensen L, McKeever SWS, Wintle AG. Optically stimulated luminescence dosimetry. Elsevier; 2003.
Contents lists available at ScienceDirect
Physica Medica journal homepage: www.elsevier.com/locate/ejmp
Original paper
Characterization of OSL dosimeters for use in dose assessment in Computed Tomography procedures Louise Giansante, Josilene C. Santos, Nancy K. Umisedo, Ricardo A. Terini, Paulo R. Costa
T
⁎
Nuclear Physics Department, Institute of Physics, University of São Paulo, São Paulo, Brazil
A R T I C L E I N F O
A B S T R A C T
Keywords: OSL Dosimetry Instrumentation Computed Tomography Radiology
This study describes the characterization of an Al2O3:C OSLD (Landauer’s Luxel™ tape) for dose evaluation in Computed Tomography. The irradiations were conducted using both a constant potential X-ray equipment and a 64-slice clinical CT scanner, and the readouts were performed using a Risø TL/OSL reader. The following aspects were studied: batch homogeneity, energy response, linearity of dose response, reproducibility, reusability, and effect of uncertainties with the normalization of OSL signals per their response to beta radiation. A group of 330 dosimeters from the 452 irradiated with the same dose presented OSL signals within the interval of 4.7% from the average. The dosimeters presented energy-dependent response in good agreement with results found in the literature. The air kerma response of the OSL signal showed a linear trend for both the constant potential X-ray device and the clinical CT scanner, with differences in their slopes of approximately 10%. Reproducibility, reusability, and effect of beta normalization were analyzed by separating 72 dosimeters in 3 groups. The results obtained in this study together with those of previous works indicate that this type of dosimeter is adequate for dose evaluation in CT clinical applications.
1. Introduction Since the development of the first Computed Tomography (CT) equipment in 1970, this diagnostic imaging modality has been rapidly expanding [1]. The radiation dose absorbed by patients due to this technique has become a concern among radiologists, researchers, and manufacturers [2–4], leading to the development of different methods to evaluate the absorbed dose in such examinations [5]. Ionization chambers (IC), thermoluminescence (TL) and, more recently, optically stimulated luminescence (OSL) dosimetry, for instance, have been widely used in order to estimate doses in vivo, in post-mortem subjects, and in phantoms. Lavoie et al. [6], for instance, proposed an experimental method to assess the nanoDot Al2O3:C dosimeter (Landauer, Inc.), aiming to apply this system in CT procedures. In the study conducted by Funama et al. [7], the authors measured dose profiles in radiosensitive organs in thorax and breast regions due to CT Coronary Angiography (CTCA) procedures. Therapeutic and diagnostic beams were assessed in the low-medium energy range by Spasic & Adam [8]. More recently, the optimal bleaching conditions were assessed with fully filled traps Al2O3:C dosimeters in order to evaluate changes in their dose sensitivity [9]. In the review proposed by Alaei & Spezi [10], the authors present
⁎
several methods used to calculate dose due to CBCT procedures, including a variety of phantoms and dosimeters. The nanoDot Al2O3:C from Landauer was also applied to measure out-of-field doses in radiotherapy, and results showed good agreement with results obtained with other types of dosimeters [11]. Takegami and colleagues [12] have used a modern 320-slice CT scanner to evaluate the entrance surface dose (ESD) in phantoms and in patients. Results reported by these authors and others indicate that this dosimetry system offers many advantages compared with other systems, such as efficiency, accuracy, linearity, and good spatial resolution, besides keeping the signal after the readouts. The present study proposes an experimental methodology for the characterization in the diagnostic range of an Al2O3:C optically stimulated dosimeter (Landauer, Inc.). The objective of the present study was to verify the applicability of this type of dosimeter in CT organ dose assessment using adult and pediatric anthropomorphic phantoms in clinical CT machines [13]. The authors demonstrate the potentiality of this technology for this kind of purpose, but emphasize that although several of the OSL properties are well documented in the literature, in special for personal dosimetry, its applicability for other purposes were not still exhaustively validated by scientific methodologies. This paper intends to contribute for this validation process.
Corresponding author at: 1371 Matão St., 05508-090 São Paulo, SP, Brazil. E-mail addresses: [email protected] (L. Giansante), [email protected] (P.R. Costa).
https://doi.org/10.1016/j.ejmp.2018.02.009 Received 4 October 2017; Received in revised form 10 February 2018; Accepted 13 February 2018 1120-1797/ © 2018 Published by Elsevier Ltd on behalf of Associazione Italiana di Fisica Medica.
Physica Medica 47 (2018) 16–22
L. Giansante et al.
2. Materials and methods
the literature [23,24]. For TL measurements each dosimeter was heated from the room temperature until 350 °C at a constant rate of 10 °C per second [15]. The TL signal was extracted by integrating the TL curve until 350 °C.
The following characteristics of OSLDs were investigated: batch homogeneity, energy response, linearity of dose response, reproducibility, reusability, and beta response normalization. The Landauer Luxel™ tape (Landauer, Inc., Glenwood, USA), from a 73 m roll measuring 20.0 mm in width and 0.3 mm in thickness, produced with carbon-doped aluminum oxide crystals converted into powder [14], was used in the present study. The tape was fractionated by the authors into disks measuring 3 or 5 mm in diameter, according to their application in the present study. A group of pre-selected thermoluminescent dosimeters (TLDs), with 3 mm × 3 mm × 1 mm, composed of LiF:Mg,Ti (TLD-100, Harshaw Chemical Company, OH, USA), was simultaneously irradiated along with the 3 mm OSLDs in a set of control measurements for comparison of results. Before each exposure, the OSLDs were optically treated (bleaching) for 8 h under fluorescent light (Digilight, ATEK, São Paulo, Brazil). Similarly, the TLDs were heated (annealing) at 400 °C for 1 h followed by 100 °C for 2 h. Both procedures aim to erase the effects of previous irradiations, stabilizing the sensitivity and the background of the dosimeters so that their properties remain invariable throughout usage [15]. Two X-ray sources were used in the present study: a high precision X-ray machine used for personal dosimeter calibration laboratory was adopted for the basic OSLD characterization tests and a clinical machine was used for linearity tests. The irradiations for characterization purposes were performed using a constant potential X-ray tube MCN 421 (Philips, Germany). RQR and RQT X-ray beam standards [16,17] were validated and characterized in this equipment [18] and used during the procedures described in the present study. A clinical 64-slice CT scanner, Brilliance 64 (Philips, Germany), from the Institute of Radiology of the School of Medicine of the University of São Paulo (InRad/FMUSP), was used in an additional set of linearity tests in order to verify the degree of similarity between the two machines. Two ionization chambers (ICs) were used. A 30 cc IC, model 233610576 (PTW, Freiburg, Germany), which is used as a reference chamber for absolute dosimetry in calibration laboratories [19], was used in the batch homogeneity and energy response tests. This chamber was coupled to an electrometer Unidos E (PTW, Freiburg, Germany) and it was chosen because of its high accuracy and signal stability. The second chamber was a 0.6 cc IC (Radcal Corporation, Monrovia, CA, USA), model 10X5-0.6. This device is currently applied for dose measurements in multi-slice Computed Tomography scanners [20,21] and it was chosen because its small volume allowed high accurate measurements in the isocenter of the clinical CT X-ray beam. It was used in the linearity, reproducibility, reusability and beta normalization effect tests. This chamber was coupled to an electrometer model 9010 (Radcal Corporation, Monrovia, CA, USA). Both ICs and electrometers were calibrated by an SSDL (Secondary Standard Dosimetry Laboratories) [16,17]. The experimental arrangement, combining X-ray source and IC, is described in the following paragraphs according to each experiment. A Risø TL/OSL reader, model DA-20 (DTU Nutech. Inc., Roskilde, Denmark), with a built-in 90Sr/90Y beta source and blue light stimulation, was used to read the information from the dosimeters. A bi-alkali photomultiplier tube (model 9235QB) was used to detect both OSL and TL signals. An ultraviolet transmitting broad-band pass filter (Hoya U340) and a blue filter pack (Schott BG-39 + BG-3) were used in front of the photomultiplier tube for OSL and TL measurements respectively [22]. Continuous blue light stimulation was applied to each disk for 90 s, and the OSL signal was extracted from the initial signal of each OSL curve. After performing the first readout, a group of OSLDs were irradiated with beta radiation for 2 s and read again. Their responses to beta radiation were used to individually normalize the OSL signals from the X-ray radiation, decreasing the uncertainties, as once suggested in
2.1. Batch homogeneity A group of 452 dosimeters was irradiated using an MCN 421 X-ray tube to evaluate the batch homogeneity of the OSLD response to the same beam quality. The RQT 9 X-ray beam quality was used, as it usually corresponds to the reference radiation quality for calibration of instruments utilized in Computed Tomography dosimetry. The 30 cc IC was positioned 5 m away from the X-ray tube, and the air kerma measured was 20.4 ± 0.3 mGy. 2.2. Energy response OSL energy response was evaluated for eight different standard Xray beam qualities: RQR 2, RQR 4, RQR 5, RQR 6, RQR 8, RQT 8, RQT 9, and RQT 10. The RQR series of radiation qualities represent the beam incident on the patient in general radiography, whereas the RQT series simulate the unattenuated beam used in CT procedures [16]. Differences between each series are due to the X-ray tube voltage and filtration and, therefore, the effective energy of the X-ray beam. Eight groups of three OSLDs were irradiated using the MCN 421 X-ray tube, and the corresponding air kerma values were measured with the 30 cc IC positioned 5 m away from the focal spot of the X-ray tube. The OSL signal of each quality, after readout, was evaluated by calculating the mean of three dosimeter responses and the uncertainties were obtained as the standard deviation of the mean (k = 1).1 The effective energy, Eeff, of each X-ray beam quality was then determined (Eq. (1)) and their values were related to the OSL signal obtained for each quality.
⎛ μ ⎞ (Eeff ) = ln(2) HVL×ρAl ⎝ ρ ⎠Al ⎜
⎟
(1)
The HVL for each X-ray beam quality was determined experimentally, based on data provided by the IAEA Technical Series Report 457 [16] and the aluminum density was determined with data provided by the National Institute of Standards and Technology (NIST) [25]. Therefore, with ρAl = 2.699 g/cm3 and values of HVL, mass attenuation
() μ
coefficient ⎛ ρ ⎞ for each X-ray beam quality were determined. The Al ⎠ ⎝ XCOM program [26] was used to find the energies associated to each
()
μ ρ Al
values, adopted as effective energy corresponding to each HVL
value. Table 1 summarizes tube voltage (kV), external filtering, HVL, and Eeff for each X-ray beam quality.
2.3. Linearity of response The linearity of the OSL responses to the incident air kerma were first assessed using the Philips 64-slice CT scanner and the Radcal 0.6 cc IC. The IC and the dosimeters were positioned in the center of the gantry on an acrylic holder attached to a support (Fig. 1). Five groups of three unused OSLDs were irradiated along with five groups of three TLDs. These experiments used a tube voltage of 120 kV and five different values of tube current-time product ranging from 30 mAs to 400 mAs, using different number of rotations. The resulting IC readings for these settings ranged from 3.9 mGy to 51.8 mGy. After these experiments, the OSLDs were read and irradiated with the beta source of the Risø reader, and their responses to beta radiation were used to individually normalize their responses to the X-ray. 1 It is reasonable to assume that the values obtained with the three dosimeters follow a Normal distribution, thus the proper t value was applied to the standard deviation of the mean for k = 1 (68.3% of confidence).
17
Physica Medica 47 (2018) 16–22
L. Giansante et al.
Table 1 X-ray beam qualities RQR and RQT series used to assess the energy response of the analyzed dosimeters. X-ray tube voltage (kV), external copper filtration (mm Cu), nominal first HVL (mm Al), and determined effective energy are presented for each radiation quality. Radiation quality
X-ray tube voltage1 [kV]
External filtration1 [mm Cu]
Nominal first HVL [mm Al]
Determined effective energy [keV]
RQR 2 RQR 4 RQR 5 RQR 6 RQR 8 RQT 8 RQT 92 RQT 10
40 60 70 80 100 100 120 150
– – – – – 0.20 0.25 0.30
1.42 2.19 2.58 3.01 3.97 6.90 8.40 10.10
25.3 29.7 31.3 33.2 37.9 49.7 56.1 64.2
Table 3 OSL groups considered for the reproducibility and reusability tests. Group
Cycle
Group 1 Group 2 Group 3
Bleaching, reading, irradiation, reading Bleaching, irradiation, reading Bleaching, irradiation, reading, irradiation with beta source, reading
generate calibration curves, which can be used, for example, to evaluate doses due to CT procedures. The techniques applied in each situation are summarized in Table 2.
2.4. Reproducibility, reusability, and beta normalization effect Reproducibility and reusability of OSLDs were evaluated using the 0.6 cc IC positioned 1 m away from the focal spot of the MCN 421 X-ray tube and X-ray beam quality RQT 9. Dosimeters were fractionated into
1
Data retrieved from the IAEA Technical Series Report 457 [16]. Reference radiation quality usually applied in calibration of equipment for CT dosimetry; therefore, data were normalized for this quality. 2
Fig. 1. Left: The ionization chamber at the isocenter of the gantry of the 64-slice Philips Brilliance 64 CT Scanner (Philips, Germany). Right: The 0.6 cc ionization chamber placed on an acrylic holder. Air kerma was measured at this position and then each group of TLD and OSLD was positioned at the same place.
disks measuring 5 mm in diameter for these experiments. A tube voltage of 120 kV, tube current of 9 mA and exposure time of 50 s were used, so that the ionization chamber reading was 20.3 ± 0.3 mGy. Four exposures were performed in a one-week period using the same parameters, with 72 dosimeters separated in three distinct groups of 24 each. Groups were separated according to Table 3, in order to choose the most accurate method. The same cycle was performed for each group four times. Reproducibility test aimed to evaluate the precision of dosimeter responses under the same irradiation conditions. Reusability test aimed to verify the efficacy of the bleaching and the possibility of reusing the dosimeters in subsequent irradiations.
Table 2 Parameters used to assess linearity of OSLD responses in distinct situations. At the laboratory, dosimeters and IC were placed 1 m away from the X-ray tube focal spot. At the hospital, dosimeters and IC were placed in the center of the gantry of the Philips Brilliance 64 CT Scanner. Source/ chamber
Tube voltage (kV)
Tube current-time product (mAs)
Air Kerma (mGy)
TLD group
OSLD group
CT Scanner/ 0.6 cc
120
30 100 150 200 400
3.92 ± 0.09 13.05 ± 0.28 19.55 ± 0.41 26.01 ± 0.56 51.83 ± 1.10
1 2 3 4 5
1 2 3 4 5
MCN 421 Xray tube/ 0.6 cc
120
90 270 540 1080 2700
4.02 ± 0.10 12.90 ± 0.31 26.25 ± 0.63 53.0 ± 1.3 133 ± 3
1 2 3 4 5
1 2 3 4 5
3. Results and discussion 3.1. Batch homogeneity The OSL responses to the X-ray irradiation presented an approximate Gaussian behavior (Fig. 2), which was verified by AndersonDarling statistics performed with the Wolfram Mathematica® 10.0 software. A group of 330 dosimeters from the 452 irradiated disks (73% of the initial batch) presented OSL signals within ± 4.7% around the average (85021 ± 4018 counts). This demonstrates that most of the dosimeters respond in a similar way under the same irradiation conditions. Given the possible differences in dosimeters shape, size, and uniformity due to their small size and the manually fractioning process, a 4.7% variation was considered small. The uniformity of the OSL Luxel™
In another set of measurements, five groups of three OSLDs and five groups of three TLDs along with the 0.6 cc IC were positioned one meter away from the focal spot of the MCN 421 X-ray tube. These experiments used a tube voltage of 120 kV, tube current of 9 mA, and five different values of exposure time ranging from 10 s to 300 s. The resulting IC readings for these settings ranged from 4 mGy to 133 mGy. These measurements aimed to verify the applicability of the CT standard qualities with the MCN 421 X-ray tube for irradiating OSLD or TLD to
18
Physica Medica 47 (2018) 16–22
L. Giansante et al.
author indicated that the aluminum oxide powder in the OSL strip was not homogeneously distributed. However, this non-uniformity did not affect the results of CT dose profiles measurements [27]. Recently, Han et al. [28] evaluated the characteristic of a 5-mm Landauer nanoDot for an 80 kV cephalometric exposure. Homogeneity was evaluated for a batch of 25 dosimeters, and the authors reported a variation of 1.67%.
3.2. Energy response The responses of the OSLDs followed a non-uniform energy dependence (Fig. 3), similar to those found in the literature [6,29–31]. Gasparian et al. [30] investigated the responses of two types of OSLDs: Al2O3:C single crystals of 5 mm in diameter and 0.9 mm in thickness and Al2O3:C Luxel dosimeters of 7 mm in diameter and 0.3 mm in thickness. Malthez et al. [31] conducted similar analysis with five different TL/OSL materials, including the Luxel OSLD cut in 7 mm of diameter and 0.3 mm in thickness. To compare the results of this study with those previously published, the authors provided their data, which was normalized to the results of this paper. The results verified by those authors, after normalization, and the ones obtained in this work presented similar energy-dependent response (Fig. 3). In CT procedures, 120 kV is the most commonly applied voltage for adult protocols [6]; however, lower tube voltages such as 80 kV, 90 kV, or 100 kV are also routinely used, depending on the clinical procedure and on the anatomy of the patient. For each case, there is a different value of HVL, which makes the evaluation of the behavior of the dosimeters necessary under these circumstances [6]. Such energy dependence in diagnostic energy range can be corrected with proper calibration.
Fig. 2. Distribution of the OSLD responses. 452 unused dosimeters were irradiated at the same time using the same technique.
3.3. Linearity of response The OSL signal dependence to the air kerma is presented in Fig. 4. An expected linear dependence was found when the 0.6 cc IC was irradiated by the MCN 421 X-ray tube and the Philips Brilliance 64 CT scanner together with TLDs and OSLDs. However, the slope of the linearity curve for the X-ray tube was 10% higher than the one obtained with the CT scanner for the TLDs and 8% higher for the OSLDs (Fig. 4). Those differences might be related, for instance, to the degree of similarity of the X-ray beam qualities of these two equipment, flux dependence, and to scattering conditions. The results of the linearity evaluation are summarized in Table 4.
Fig. 3. OSLD energy responses obtained in this study for eight different X-ray beam qualities. The mean OSL signal obtained for each quality was normalized to the one verified for the RQT 9 X-ray beam quality. The other two symbols correspond to the energy responses for the Al2O3:C Luxel dosimeters obtained in distinct works. These data were normalized to the data found in this work for comparison.
tape was investigated by Gasparian [27]. The author performed this analysis in OSL strips cut from the original tape irradiated using a linear accelerator with photon energy of 6 MeV. Results reported by this
Fig. 4. Comparison of the linearity responses of the OSLDs (left) and TLDs (right) to the air kerma measured by a 0.6 cc IC in the MCN 421 X-ray tube and in the CT scanner. The OSL signal was normalized by their responses to beta radiation, as previously explained (Section 2.3).
19
Physica Medica 47 (2018) 16–22
L. Giansante et al.
Table 4 OSLD and TLD linearity response parameters obtained for distinct situations for 120 kV. Source
Ionization chamber (IC)
Dosimeter
Distance source-IC/dosimeter [m]
Slope (counts/mGy)
R2
X-ray tube CT scanner X-ray tube CT scanner
0.6 cc 0.6 cc 0.6 cc 0.6 cc
OSLD OSLD TLD TLD
1 Center of the gantry 1 Center of the gantry
0.141 ± 0.0021 0.135 ± 0.0021 (73.5 ± 1.8) × 103 (66.2 ± 0.9)8103
0.9986 0.9990 0.9992 0.9975
1
OSL signal normalized to beta radiation.
Fig. 5. Dosimeter responses per irradiation for groups 1 and 2 and normalized OSL signal for group 3. For each data set, whiskers present the variation range (minimum and maximum values), the box contains 50% of data (75%–25%), the horizontal line inside the box represents the median value of the responses and the square represents the mean value of the data. Groups 1 and 2 presented outliers.
respectively, 2.4% and 0.26%. The comparative evaluation between the results obtained with dosimeters from Groups 1, 2, and 3 highlights the effect of normalizing each OSL signal (SX-ray) by their signal obtained after the irradiation with a beta source (Sβ). With the “normalized OSL signal”, dependence with the homogeneity of the crystals in the dosimeter, caused by the manual process of fractionating the tape, or stability of the reader, for instance, are taken into account and individually corrected. In Fig. 5, dosimeter responses from three groups are presented in box plots per irradiation. The absence of outliers in the responses from Group 3 highlights the efficacy of individually normalizing the OSL signals per their responses to beta radiation. Dosimeter responses from Groups 1 and 2 presented outliers in all readings.
3.4. Reproducibility, reusability, and beta normalization effect After irradiating the three groups under the same conditions, mean, standard deviation and coefficient of variation (CV) for each dosimeter group were recorded. OSL signals from each group were normalized so that their CVs could be comparable. This study indicated a variation among irradiations from 3% to 5% for Group 1, from 4% to 5% for Group 2 and from 1% to 2% for Group 3. Similarly, during the process of irradiating and erasing the 24 dosimeters from the three groups, each dosimeter was identified and their signal was individually recorded along the readings to assess the reusability. After performing 4 cycles (Table 3), CV values varied from 1% to 4% for dosimeters from Groups 1 and 2 and from 1% to 2% for dosimeters from Group 3. In a similar analysis, Musa and colleagues [32] measured the CV of ten repeated measurements using two nanoDots. In that study, the authors used a higher dose (10 and 1 Gy), and the CV values encountered were, 20
Physica Medica 47 (2018) 16–22
L. Giansante et al.
and Development) for the financial support (process 132554/2015-1), and the Institute of Radiology of the School of Medicine of the University of Sao Paulo (InRad/FMUSP) for the partnership (CAAE 55420616.3.0000.0068). We thank Malthez and colleagues, and Gasparian and colleagues for providing their data that we used for comparison. We also thank the CNPq/FAPESP for funding the project by INCT – Metrology of Ionizing Radiation in Medicine (process 2008/ 57863-2).
4. Conclusions This study presented a methodology for the characterization of the Al2O3:C Luxel OSLD tape manufactured by Landauer, Inc., manually fractioned for CT applications. The results of some tests were compared with those obtained with the LiF:Mg,Ti thermoluminescent dosimeters and with the results found in the literature regarding works with similar dosimeters. The dosimeters responses were analyzed for homogeneity, linearity with the incident air kerma, reproducibility, reusability, and energy-dependence to distinct effective energies. The batch homogeneity presented a Gaussian behavior with 4.7% of standard deviation, which demonstrates that most of the dosimeters in the same batch respond in a similar way. However, the normalization of the OSL signals to their responses to beta radiation proved to be efficient in decreasing uncertainties related to differences among dosimeters and by non-stability of the reader. The coefficient of variation of the responses decreased by at least 3%, thus proving the efficiency of such analysis. Reproducibility and reusability tests conducted in this study demonstrated that this kind of dosimeter can be reused in subsequent radiations for low-dose procedures. There were no relevant differences in dosimeter responses after 4 consecutive irradiations and the 8-h bleaching process proved to be efficient. When working with higher doses, however, a more extensive characterization might be needed to ensure doses are properly erased. Moreover, individual readings of each dosimeter can present a higher variation, thus it is recommended to use more than one dosimeter per exposure, and to include the standard deviation of their mean as a component of the uncertainty for a more reliable result. Linearity was assessed up to 140 mGy with a 0.6 cc IC, a constant potential X-ray tube, and a 64-slice CT scanner. This is sufficient for CT procedures, but a more extensive range of characterization is needed for other clinical applications with higher doses. Moreover, the linear response indicates that these curves can be applied as calibration curves to convert dosimeter readings into air kerma values. These experiments aimed to evaluate the applicability of the CT standard qualities with the MCN 421 X-ray tube to generate calibration curves. The RQT standard qualities are obtained to reach a specific nominal HVL by adding an external filtration to simulate the non-attenuated beam used in CT procedures [16]. If the X-ray tube and the CT scanner beam qualities were the same, one could not expect discrepancies observed in Fig. 4. Thus, from the obtained results, one can infer that it is preferable to calibrate dosimeters with the same equipment used in the measurements. The energy-dependent response of this kind of dosimeter can be explained in terms of the effective atomic number (Zeff) of Al2O3:C, which is 11.28 [33]. It means that the total photoelectric effect crosssection of this material is higher than that of materials such as soft tissue, air, or water, especially for the low-energy photons typically found in radiology [31]. For such radiation, the energy deposited in the dosimeter is higher than the energy deposited in biological tissues for the same X-ray fluence [31,33]. As a consequence, these dosimeters present overresponse to low-energy x-rays, typically found in radiology [34,35]. Finally, some advantages of these dosimeters over the TLD can be highlighted. The OSLD optical readout process is simpler than the TLD heating readout process, because it eliminates the need of a reliable and reproducible heating system [36]. Furthermore, considering that Landauer provided this material as a tape, the user can fractionate it according to the desired application. In a subsequent work this type of dosimeter was applied in anthropomorphic phantoms to evaluate organ doses due to CT examinations as an alternative to the TLDs [13].
References [1] Rehani MM, Bongartz G, Kalender W, Golding SJ, Gordon L, Murakami T, et al. Managing patient dose in computed tomography. Ann ICRP 2000;30:7–45. [2] Brenner DJ, Hall EJ. Computed tomography – an increasing source of radiation exposure. N Engl J Med 2007;357:2277–84. http://dx.doi.org/10.1056/ NEJMra072149. [3] IAEA. Dosimetry in diagnostic radiology for paediatric patients. Vienna: International Atomic Energy Agency (IAEA Human Health Series No. 24); 2014. [4] Paul J, Chacko A, Saccomandi P, Vogl TJ, Nour-Eldin N-EA. Effect of diagnostic cone-beam computed tomography protocols on image quality, patient dose, and lesion detection. Phys Med 2016;32:1575–83. http://dx.doi.org/10.1016/j.ejmp. 2016.11.111. [5] Longo M, Genovese E, Orlandi C, Donatiello S, Cassano B, Colajacomo M, et al. A novel method for CT dosimetry with a suspended phantom setup. Phys Med 2017;40:122–9. http://dx.doi.org/10.1016/j.ejmp.2017.07.006. [6] Lavoie L, Ghita M, Brateman L, Arreola M. Characterization of a commerciallyavailable, optically-stimulated luminescent dosimetry system for use in computed tomography. Health Phys 2011;101:299–310. http://dx.doi.org/10.1097/HP. 0b013e31820f8e0e. [7] Funama Y, Taguchi K, Utsunomiya D, Oda S, Murasaki H, Yamashita Y, et al. Dose profiles for lung and breast regions at prospective and retrospective CT coronary angiography using optically stimulated luminescence dosimeters on a 64-detector CT scanner. Phys Med 2012;28:76–82. [8] Spasic E, Adam JF. Optically stimulated luminescence for diagnostic and therapeutic low to medium energy X-ray beams experimental dosimetry. Phys Med 2013;29:e42. [9] Park S-Y, Choi CH, Park JM, Chun M, Han JH, Kim J. Sensitivity and stability of optically stimulated luminescence dosimeters with filled deep electron/hole traps under pre-irradiation and bleaching conditions. Phys Med 2017;38:81–7. [10] Alaei P, Spezi E. Imaging dose from cone beam computed tomography in radiation therapy. Phys Med 2015;31:647–58. http://dx.doi.org/10.1016/J.EJMP.2015.06. 003. [11] Harrison R. Out-of-field doses in radiotherapy: input to epidemiological studies and dose-risk models. Phys Med 2017;42:239–46. [12] Takegami K, Hayashi H, Yamada K, Mihara Y, Kimoto N, Kanazawa Y, et al. Entrance surface dose measurements using a small OSL dosimeter with a computed tomography scanner having 320 rows of detectors. Radiol Phys Technol 2017;10:49–59. http://dx.doi.org/10.1007/s12194-016-0366-1. [13] Giansante L. Organ doses evaluation for computed tomography procedures using TL and OSL dosimeters. University of São Paulo; 2017. [14] Yukihara EG, McKeever SWS. Optically stimulated luminescence (OSL) dosimetry in medicine. Phys Med Biol 2008;53:R351–79. http://dx.doi.org/10.1088/00319155/53/20/R01. [15] Furetta C. Questions and answers on thermoluminescence (TL) and optically stimulated luminescence (OSL). World Scientific; 2008. [16] IAEA. Dosimetry in diagnostic radiology: an international code of practice. Vienna: International Atomic Energy Agency (Technical Report Series No.457); 2007. [17] IEC. Medical diagnostic X-ray equipment – radiation conditions for use in the determination of characteristics (1994 Standard IECS No. 61267). Int Electrotech Comm 2005. [18] Burger A, Costa PR. Implementação e Validação de Qualidades de Feixes de Raios X Segundo o Código de Prática da IAEA. 2012. [19] PTW. Cylinder stem ionization chamber 23361. http://www.ptw.de/cylinder_stem_ chamber.html; 2017. 52. [20] AAPM. Comprehensive methodology for the evaluation of radiation dose in X-ray computed tomography AAPM task group 111 (AAPM report 111). Coll Park, MD: Am Assoc Phys Med; 2010. [21] Radcal. 10 × 5 series ion chambers – technical specifications 2013; 2017. [22] Nutech DTU. Guide to the Risø TL/OSL reader. DTU Nutech; 2015. [23] Yukihara EG, Yoshimura EM, Lindstrom TD, Ahmad S, Taylor KK, Mardirossian G. High-precision dosimetry for radiotherapy using the optically stimulated luminescence technique and thin Al2O3:C dosimeters. Phys Med Biol 2005;50:5619–28. http://dx.doi.org/10.1088/0031-9155/50/23/014. [24] Yukihara EG, McKeever SWS. Optically stimulated luminescence: fundamentals and applications. John Wiley & Sons; 2011. [25] NIST. XCOM: photon cross sections database 1998. 2017. [26] Berger MJ, Hubbell JH, Seltzer SM, Chang J, Coursey JS, Sukumar R, et al. NIST standard reference database 8 (XGAM). XCOM phot cross sect database; 2008. [27] Gasparian PBR. Methodological developments for application of optically stimulated luminescence (OSL) in medical dosimetry. Oklahoma State University; 2009. [28] Han SC, Kim KB, Jung H, Ji Y, Park S. Assessment of diagnostic multileaf collimator for cephalometric exposure reduction using optically stimulated luminescent
Acknowledgments The authors thank CNPq (Brazilian National Council for Research 21
Physica Medica 47 (2018) 16–22
L. Giansante et al.
[29] [30]
[31]
[32]
dosemeters. Radiat Prot Dosimetry 2016;174:102–8. http://dx.doi.org/10.1093/ rpd/ncw080. Al-Senan RM, Hatab MR. Characteristics of an OSLD in the diagnostic energy range. Med Phys 2011;38:4396–405. http://dx.doi.org/10.1118/1.3602456. Gasparian PBR, Vanhavere F, Yukihara EG. Evaluating the influence of experimental conditions on the photon energy response of Al2O3:C optically stimulated luminescence detectors. Radiat Meas 2012;47:243–9. http://dx.doi.org/10.1016/j. radmeas.2012.01.012. Malthez ALMC, Freitas MB, Yoshimura EM, Button VLSN. Experimental photon energy response of different dosimetric materials for a dual detector system combining thermoluminescence and optically stimulated luminescence. Radiat Meas 2014;71:133–8. http://dx.doi.org/10.1016/j.radmeas.2014.07.018. Musa Y, Hashim S, Karim MKA, Bakar KA, Ang WC, Salehhon N. Response of optically stimulated luminescence dosimeters subjected to X-rays in diagnostic energy
[33]
[34]
[35]
[36]
22
range. J Phys Conf Ser 2017;851:12001. http://dx.doi.org/10.1088/1742-6596/ 851/1/012001. Bos AJJ. High sensitivity thermoluminescence dosimetry. Nucl Instrum Methods Phys Res Sect B 2001;184:3–28. http://dx.doi.org/10.1016/S0168-583X(01) 00717-0. Mobit P, Agyingi E, Sandison G. Comparison of the energy-response factor of LiF and Al2O3 in radiotherapy beams. Radiat Prot Dosimetry 2006;119:497–9. http:// dx.doi.org/10.1093/rpd/nci676. Perks CA, Yahnke C, Million M. Medical dosimetry using optically stimulated luminescence dots and microStar readers. Proceedings of IRPA12: 12. Congress of the International Radiation Protection Association: Strengthening Radiation Protection Worldwide - Highlights. Global Perspective and Future Trends; 2010. Bøtter-Jensen L, McKeever SWS, Wintle AG. Optically stimulated luminescence dosimetry. Elsevier; 2003.