{"\ufeffDOSIMETRIC CHARACTERISATION OF THE NANODOT OPTICALLY STIMULATED LUMINESCENT DOSIMETER FOR USE IN NATIONAL ELECTRON BEAM DOSIMETRY AUDIT SERVICES FOR RADIOTHERAPY FACILITIES\n\nN. Abdullah1,2,*, N. Mohd Noor1,3,*, Z. Kamarul Zaman4, M. Mohammad Zahid5, N.M Ung6\n\n1Medical Physics Laboratory, Department of Radiology, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia\n2Medical Physics Laboratory, Radiation Metrology Group, Malaysian Nuclear Agency, 43000 Kajang, Selangor, Malaysia\n3Medical Physics Unit, Hospital Sultan Abdul Aziz Shah, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia\n4Department of Medical Physics, University of Malaya Medical Centre, 59100 Kuala Lumpur, Malaysia\n5Department of Radiotherapy and Oncology, National Cancer Institute, 62250 WP Putrajaya, Malaysia\n6Dean Office, Faculty of Medicine, University of Malaya, 59100 Kuala Lumpur, Malaysia\n\n*Corresponding Authors\u2019 Email: 1,2hayatie@nm.gov.my": null, " 1,3noramaliza@upm.edu.my\nArticle History: Received April 24, 2024": null, " Revised May 14, 2024": null, " Accepted May 17, 2024\n\nABSTRACT: The Malaysian Nuclear Agency's secondary standard dosimetry laboratory (SSDL) aims to establish a national dosimetry audit service for radiotherapy facilities. For this purpose, a nanoDot optically stimulated luminescent dosimeter (OSLD) was selected as the transfer dosimeter for the audit program. The study aims to establish the basic dosimetric characteristics and associated correction factors of nanoDot OSLD for use in electron beam dosimetry audits. An investigation of the dosimetric characteristics of the nanoDot, comprising the sensitivity correction factor (SCF), dose-response linearity, beam energy dependency, signal depletion per readout, and signal fading when subjected to electron beams, was conducted. A preliminary electron beam dosimetry audit using nanoDot OSLD was performed for two radiotherapy facilities under both reference and non-reference conditions. The measurement uncertainty of the absorbed dose for the nanoDot OSLD was also estimated. The mean SCF of the 91 nanoDot OSLD was 1.001 \u00b1 0.25%. The dose-response curves for the 6 MeV and 9 MeV beams exhibited linear characteristics, with a determination coefficient of 0.9982 for the dose range of 50\u2013300 cGy. However, a high energy dependency was observed at 12 MeV, resulting in a deviation of 4.08% compared to that at 6 MeV. The nanoDot signal decreased by 0.03% after 100 readouts and faded by 3.20% at 70 days post-irradiation. It is noteworthy that all audit results from the six electron beams were in compliance with the tolerance limit of \u00b1 5%, with mean dose deviations of -1.66% \u00b1 0.81% and -1.37% \u00b1 0.65% for the reference and non-reference conditions, respectively. The combined uncertainty was estimated to be \u00b1 1.41% (coverage factor, k ": "1). National electron beam dosimetry audits using nanoDot OSLD can now be implemented as a regular service.\n\nKEYWORDS: radiotherapy dosimetry audit", " electron beam": null, " optically stimulated luminescent dosimeter": null, " nanoDot OSLD.\n\n1.0 INTRODUCTION\nSafe and effective radiotherapy relies on the accuracy of dose delivery to the target volume in the patient, which is typically within \u00b1 5% of the prescribed dose at a 95% confidence level, as recommended by the International Commission on Radiation Units and Measurements (ICRU)[1]. Any errors in radiotherapy dosimetry can lead to radiation injuries and severe complications [2]. To prevent such errors, dosimetry audits are conducted at the national or international level as part of the quality assurance program (QAP) in radiotherapy. These audits have been successful in identifying errors, providing support for identifying the sources of errors, and rectifying them [3], [4]. The audit also serves as an early error detection mechanism, which is essential for taking prompt corrective action to safeguard patients from potential harm. This practice will improve dosimetry practices and reduce the likelihood of errors occurring, ultimately affecting patient health [5]. According to D. Van Der Merve [6], an independent dosimetry audit should be performed for every new installation and regularly. This is crucial because dosimetry audits provide medical physicists with confidence in applying new radiotherapy modalities and techniques [7]. \nIn keeping with the International Atomic Energy Agency (IAEA) guidelines, Malaysian radiotherapy facilities have actively participated in IAEA/World Health Organization (WHO) postal radiotherapy dosimetry audits since 2011, with a focus on assessing the absorbed dose of water from photon beams under reference conditions [8]. Presently, there are 35 radiotherapy centres in the country, comprising seven government hospitals and 28 private facilities [9]. One government-funded radiotherapy service was provided through a contract with a private institution. In total, 93 radiotherapy modalities are available, including 57 medical linear accelerators (linac), 19 brachytherapy, 7 intra-operative radiotherapy (IORT), 5 tomotherapy, 3 gamma knife, and 2 cyberknife [10]. On average, seven radiotherapy centres in Malaysia participate in the IAEA audit annually, with the highest participation reaching 12 centres in 2022. The IAEA audit results from 2011 to 2022 found that out of 202 beams checked, the majority (93%) were satisfied with an acceptance limit of \u00b15%, except for 13 photon beams (6%) and three electron beams (2%) [11]. Despite the low failure rate, this situation poses a severe risk of radiation injuries and, in extreme cases, death if not addressed promptly. Until recently, no national remote dosimetry audits had been conducted in Malaysia, as a national dosimetry audit network (DAN) has not yet been established. An international review of remote dosimetry audits indicated that electron beams are more susceptible to errors than photon beams [12], [13]. In 2021, the International Atomic Energy Agency (IAEA) initiated the Electron Audit Service for its member states, with one facility from Malaysia participating [14]. Therefore, establishing a national dosimetry audit for electron beams is crucial because of insufficient access to radiotherapy centres for IAEA audits, as priority is given to new linac installations.\nGenerally, electron remote dosimetry audits are limited to the measurement of the electron beam output under reference conditions [15], [16], with some additional parameters tested for on-site audits [17], [18], [19]. For this purpose, various detectors have been employed, including alanine, ionisation chambers, and radio photoluminescent glass dosimeters (RPLD). Each dosimeter has exclusive advantages": null, " however, the nanoDot optically stimulated luminescent dosimeter (OSLD), one of the most versatile dosimetry systems with great dosimetric characteristics and convenience of use for a large-scale audit, is preferable for remote dosimetry audits [20], [21]. Several studies have described the dosimetric characteristics of nanoDot OSLD for radiotherapy dosimetry applications [22], [23], [24], [25]. However, this study is the first to present the establishment of dosimetric characteristics and their associated correction factors for nanoDot OSLDs for remote radiotherapy dosimetry audits using electron beams for reference and non-reference conditions. These characteristics include (i) dosimeter sensitivity, (ii) dose-response linearity, (iii) beam energy dependency, (iv) signal depletion, (v) signal fading, and (vi) readout reproducibility. This was followed by the fabrication and testing of the newly fabricated nanoDot OSLD holder for electron beam audit, implementation of a preliminary dosimetry audit for electron beams in reference and non-reference conditions, and relevant measurement uncertainty of the absorbed dose from nanoDot OSLD. \n\n2.0 METHODOLOGY\n1 \n2 \n2.1 NanoDot optically stimulated luminescence dosimetry system\nThis study utilized a nanoDot optically stimulated luminescence (OSL) dosimetry system procured from Landauer Inc. (Gleenwood, USA). The nanoDot OSL material was made of Aluminum Oxide doped with Carbon (Al2O3:C) and has a thickness of 0.12 cm and a diameter of 0.5 cm. It is encased in a 1 cm \u00d7 1 cm \u00d7 0.18 cm light-tight plastic case with a mass density of 1.03 g/cm3 to protect it from light-induced signal fading. The OSL material in the disc can easily slide out of its plastic casing during the readout and optical bleaching processes. The nanoDot was read using an InLight MicroStar system installed with MicroStar software version 4.3. During the readout process, the OSL material was exposed to green light (540 nm wavelength). This process triggered the dosimeter to emit blue light with a wavelength of 430 nm, and the light signals were counted using a photomultiplier tube (PMT). The OSL signals were then converted to the actual absorbed dose in Gray by multiplying with the relevant correction factors.\nTo minimise errors due to accumulative background signals, pre-irradiation OSL signals were recorded by reading the nanoDot within one day before irradiation. Between the irradiation and readout periods, the nanoDots were kept in a closed cabinet at room temperature to minimise sensitivity and optical fading [26]. The nanoDots can be read 10 min post-irradiation to allow stabilisation of the OSL signals [27], [28]": null, " however, in this study, the irradiated nanoDots were read not earlier than 24 h after irradiation to acquire post-irradiation OSL signals. When necessary, the net OSL signal was calculated by subtracting the pre- and post-irradiation signals. At least three nanoDots were used for all measurements, which were read more than five times sequentially to provide reliable mean values and a smaller margin of error.\n\n2.2 Optical annealer\nThe OSL signals were bleached using an light emission diode (LED) X-ray illuminator (MST-4000, Minston, China). The annealer was made of a 20 W LED light panel with a luminance of 5500 cd/m2. This optical bleaching process requires manual sliding of the OSL disc out of the plastic case. The OSL disc was continuously exposed to light from the optical annealer for at least three days until the residual signal was nearly identical to the background signals (< 200 counts). Background subtraction was not performed if the background signals were minimal compared to the OSL signals (> 100,000 counts).\n\n2.3 Electron beam irradiation using a linac\nA linac Novalis Tx linear accelerator (Varian Medical Systems, Palo Alto, CA, USA) at the University Malaya Medical Centre (UMMC) was used to establish the dosimetric characteristics and relevant correction factors of the nanoDot subject to electron beams. To minimise the air gap during irradiation, the nanoDot was placed inside a polymethyl methacrylate (PMMA) slab phantom with dimensions of 30 cm \u00d7 30 cm \u00d7 0.7 cm that was designed with a slot to accommodate the nanoDot tightly. A 10 cm thick solid water phantom (type 457, Gammex RMI, USA) with dimensions of 30 cm \u00d7 30 cm was placed below the PMMA slab phantom to provide a full-scatter condition. The nanoDots were irradiated at the intended dose, delivered at a rate of 400 cGy/min at 100 cm source-surface distance (SSD) with a 10 cm \u00d7 10 cm field size at scaled depth in a PMMA slab phantom,  for 6 and 9 MeV electron beams. Considering the density of the PMMA slab phantom,  of 1.19 g/cm3 and depth scaling factor,  of 0.941 [29], the  and measurement depth equivalent in water,  was estimated as presented in Table 1. These experimental setups were the same for all dosimetric measurements unless otherwise mentioned.\nTable 1: Measurement setup of nanoDot in PMMA slab phantom for electron beam irradiation.\nBeam energy (MeV)\n6\n9\nBeam quality,  (cm)\n2.47\n3.68\nReference depth in water, Zref (g/cm2)\n1.38\n2.11\nMeasurement depth in PMMA,  (cm)\n1.05\n1.50\nScaled depth in PMMA,  (g/cm2)\n1.25\n1.79\nMeasurement depth equivalent in water,  (g/cm2)\n1.18\n1.68\nPercentage depth dose, PDD (%)\n99.65\n99.36\nTo determine the calibration coefficient, the nanodots were calibrated in terms of the absorbed dose to water under reference conditions in 6 and 9 MeV beams. A calibrated 0.4 cm3 plane parallel ionisation chamber, type PPC40 (IBA Dosimetry GmbH, Germany), connected to a PTW Unidos E electrometer, type T10009 (PTW Freiburg, Germany), was used to measure the absorbed dose to water. The absorbed dose to water was determined according to IAEA\u2019s Technical Report Series (TRS) No. 398 [29]. Detailed procedures are discussed in Abdullah et al.(2023). The calibration coefficient of the nanoDot OSLD dosimetry system was calculated as the ratio of the absorbed dose in water measured using an ionisation chamber, in cGy, and the corrected OSL readings of the nanoDot, in nC.\n\n2.4 Dosimetric characteristics and relevant correction factors\n2.4.1 NanoDot sensitivity correction factor\nThe inhomogeneous composition of the OSL material in the nanoDot produced variability in the sensitivity required for the individual sensitivity correction factor (SCF). In this study, 93 nanoDots were examined for their SCFs. Before irradiation, the pre-irradiation signal (initial background) was measured for each nanoDot. The nanoDots were then irradiated with 50 cGy under conditions of 20 cm \u00d7 20 cm field size (FS) and 100 cm SSD for a 6 MeV electron beam. A 0.5 cm bolus and a 1.0 cm slab phantom were placed above the nanoDots to provide a flat beam at the reference depth of 1.5 cm. After removing the outliers (nanoDots whose background corrected signal was outside three times the standard deviation), the SCF was calculated by taking the ratio of the average net OSL signal of all nanoDots and the net OSL signal of each nanoDot. The data were further assessed using a one-sample t-test using IBM SPSS Statistics Version 29.\n\n2.4.2 Linearity of dose with OSL signal\nTo investigate the OSL signal response as a dose function, three nanoDots were irradiated for each planned dose within 50\u2013300 cGy at 25 cGy intervals for both 6 MeV and 9 MeV beams. The linearity curves of the OSL signals against dose were plotted with linear functions fitted to the data to obtain the determination coefficients (R2). The corresponding dose-response linearity correction factor,  for each beam energy was calculated as the ratio of the OSL signal at 100 cGy to other doses. A graph of versus dose was plotted, with the linear functions fitted to the data to obtain a linear equation model. Finally, simple linear regression statistical tests were conducted to predict the value of  based on the dose for each beam energy.\n\n2.4.3 Beam energy dependency\nThe beam energy dependency of nanoDots was quantified for the most commonly used radiotherapy treatment beams: 6, 9, and 12 MeV. The nanoDots were irradiated under the reference conditions using a fabricated PMMA OSLD holder (Figure 1) following TRS-398 [29]. The energy correction factors,  were subsequently determined based on the ratio of the OSL signal emitted by the nanoDots in a 6 MeV beam relative to the other beams.\n\n2.4.4 Signal depletion per readout\nThe OSL signal in the nanoDot could be read repeatedly, but with partial signal loss. To study the signal depletion per readout, the nanoDots were exposed to 200 cGy with 6 MeV and 9 MeV beams. Without repositioning the nanoDot in the reader, signal depletion of nanoDot was observed by reading the nanoDots 100 times successively with a 10-second reading cycle. A graph of the signal depletion versus the sequential reading number was plotted, and a linear function was fitted to the data. \n\n2.4.5 Signal fading over time\nThe decay in the OSL signal as a function of time post-irradiation was assessed by exposing the nanoDots to 200 cGy with 6 MeV and 9 MeV beams. A total of 17 nanoDots were prepared, of which 15 nanoDots were exposed to the corresponding beams, and the remaining dosimeters were used for background radiation monitoring. The first OSL reading was taken 24 h after irradiation to allow for decay of the phosphorescence signals observed immediately after irradiation [31]. The irradiated nanoDots were then read once per week for ten weeks. Simultaneously, two control nanoDots were read to monitor the accumulated background radiation. Each data point was corrected using a signal depletion correction factor and accumulated background radiation. A graph of the normalised OSL signal against days post-irradiation was plotted. A logarithmic function was fitted to the data, and the standard uncertainty was calculated. \n\n2.4.6 Reproducibility of OSL readout and dosimeter\nThe reproducibility of the OSL readout was evaluated by randomly taking five nanoDots and delivering them at 200 cGy with 6 MeV and 9 MeV beams. The standard uncertainty provided by each dosimeter for various numbers of readings was compared to determine the readout reproducibility. In this study, the nanoDots were reused multiple times after bleaching as opposed to a single use. Therefore, the dosimeter reproducibility after long-term reuse of nanoDots was evaluated. To test this, the SCF for all nanoDots, except for four nanoDots used for the long-term stability of the OSL reader, was determined again following the method described in 2.4.1. A paired sample t-test was used to evaluate the differences in the data of old and new SCF for statistical significance. \n\n2.4.7 Fabrication and test of nanoDot OSLD holder for remote dosimetry audit\nThe OSLD holder for the electron beam dosimetry audit was fabricated using PMMA with a density of 1.190 g/cm3. The complete set of holders consisted of a stand, nanoDot OSLD disc, rod spacers, ring spacers, and screws (Figure 1). The stand was fabricated following the IAEA TLD standard stand [32] with a lead base to provide weight in water. The OSLD disc was designed with a 10 mm \u00d7 10 mm \u00d7 2 mm groove to fit into a single nanoDot and a watertight lid. Various ring spacer thicknesses (1, 2, and 10 mm) were used to adjust the irradiation depth of the nanoDots to correspond to the beam energy used. Two metal screws were used to open the nanodot disc lid. The scatter influence of the fabricated holder was investigated experimentally using the EBT 3 film by comparing the results to the IAEA standard holder for beams with energies of 6, 9, and 12 MeV. \n \n2.5 OSL reader stability\nThe OSL reader was warmed for 30 min to ensure system stability before use. Following this, the readers' performance was monitored according to the established measurement standards. The measurement includes dark counts from the PMT tube (DRK), count calibration using a built-in Carbon-14 (14C) radioactive source (CAL), and beam intensity from the LED. The measurement standard results were checked to ensure that they were within the specified limits for the DRK (less than 30 counts), CAL, and LED (within \u00b1 10% of the average value). Additionally, a quality control (QC) test using standard nanoDots irradiated with a Srontium-90 beta source (90Sr) was performed to determine the long-term stability of the reader.\n\n\nFigure 1: The PMMA OSLD holder for the electron beam audit. It consists of (A) a stand": null, " (B) an nanoDot OSLD disc": null, " (C) rod and ring spacers": null, " (D) screws.\n\n2.6 Preliminary dosimetry audit of electron beams\nThe primary objective of the dosimetry audits was to assess the accuracy of the absorbed dose delivered by the linac for electron beams under both reference and non-reference conditions. The audits involved six electron beams produced from four linacs at two radiotherapy centres. Prior to the audit, each centre received a set of instructions and materials including the irradiation procedure and form, a specified quantity of nanoDots for each beam (eight for irradiation and one for control), and a fabricated OSLD holder set. The centres were requested to irradiate the nanoDot in a water phantom using an OSLD holder at an absorbed dose of 100 cGy for the following conditions: (i) reference condition as defined by TRS 398, and (ii) non-reference condition at the beam's central axis with FS of 6 cm \u00d7 6 cm, 10 cm \u00d7 10 cm, and 15 cm \u00d7 20 cm at a depth of maximum dose,  and SSD greater than 105 cm.\nThe measured absorbed dose ( was calculated from the OSL signals using the following equation:\n           (1)\nwhere  is the mean of the net OSL signals from the two nanoDots,  is the nanoDot sensitivity correction factor,  is the dosimetry system calibration coefficient of 0.001712 cGy \u00b1 0.82%,  is the dose-response linearity correction factor,  is the energy correction factor,  is the fading correction factor, and  is the holder correction factor.\nThe audit results were expressed as the percentage deviation between the dose delivered by the radiotherapy centres and the absorbed dose measured from the nanoDots. \n\n2.7 Estimation of measurement uncertainty\nIn accordance with Equation 1, an uncertainty analysis of the measured absorbed dose from nanoDots was carried out utilising the guidelines outlined in the \"Guide to the expression of uncertainty in measurement\" [33]. The sources of uncertainty and the corresponding numerical values of the random (Type A) and systematic (Type B) uncertainties were determined. The total combined standard uncertainty was calculated by summing the Type A and Type B uncertainties using a quadratic method.\n\n\n3.0 RESULTS AND DISCUSSION\n\n2.8 Establishment of dosimetric characteristics and correction factors\n2.8.1 Sensitivity correction factor of nanoDot\nFigure 2 shows the distribution of the SCFs of nanoDots subjected to a 6 MeV electron beam. After eliminating the two outliers, the SCFs ranged from 0.946 to 1.060, with a mean of 1.001 \u00b1 0.25%. The nanoDots used in this study were obtained from the manufacturer and were pre-screened to \u00b1 5% uniformity of sensitivity": null, " however, the results revealed that 86 (94%) nanoDots were within this range. These findings are comparable with the published data by Retna Ponmalar et al. (2017), who reported the SCF distribution between 0.90 and 1.07, with 90% of the 200 nanoDots falling within that acceptance limit. Another study reported that 97% of 1000 nanoDots were within the acceptable limit, and the SCF distribution was between 0.930 and 1.134 [22]. These results help to justify that instead of using the SCFs provided by the manufacturer that were irradiated with Cesium-137 (137Cs), users should determine the SCF experimentally for their applications. To minimise the uncertainty due to SCFs, selected nanoDots with sensitivities of \u00b1 5% (ICRU 24, 1976) and \u00b1 3% were subsequently utilised in the dosimetric characteristic study and electron beam dosimetry audit. Five nanoDots with SCF outside the acceptance limit were used for background radiation monitoring. Further analysis using a one-sample t-test was performed to assess whether the mean SCF in these nanoDots differed from the normal SCF, which was defined as 1.000. The assumption that the SCFs were normally distributed was met, as assessed by the Shapiro-Wilk test (p ": "0.811). The mean SCF of 1.001, with a standard deviation of 0.024, was slightly higher than the normal SCF of 1.000", " however, the difference was not statistically significant (t(90) ": null, " t(88) ": null}