PHITS Simulation Analysis of the Performance of Pediatric Thyroid Screening Measurements Using NaI(Tl) Spectrometer
Article information
Abstract
Background
A NaI(Tl) survey meter was used in the pediatric thyroid screening conducted after the Fukushima Daiichi Nuclear Power Plant accident. However, this measurement method has the weakness that it is difficult to selectively identify 131I. In this study, we analyzed the performance of an energy-analyzable NaI(Tl) spectrometer using the Particle and Heavy Ion Transport code System (PHITS), which allows Monte Carlo simulation of radiation transport.
Materials and Methods
The spectrum of energy emitted by the NaI(Tl) spectrometer was simulated for a total energy absorption peak by adjusting factors that affect pulse wave height. From these simulation results, the detection limits of the NaI(Tl) spectrometer were obtained using Monte Carlo simulation.
Results and Discussion
The energy spectrum results were reproduced with an accuracy of 0.1% to 44.0% for total energy absorption peaks. The calculated detection limit for 131I activity equivalent to 100 mSv in the thyroid under a 0.2 μSv·hr−1 ambient dose rate was approximately 80–90 Bq, which could be detected for up to 38 days after 131I intake in a 1-year-old child.
Conclusion
This study demonstrated that pediatric thyroid screening using an NaI(Tl) spectrometer can practically provide greater accuracy than NaI(Tl) survey meters.
Introduction
The Great East Japan Earthquake of March 2011 caused a major accident at the Fukushima Daiichi Nuclear Power Plant (FDNPP). As a result, radioactive materials such as 131I and 137Cs were released into the environment [1]. These radioactive materials formed a plume that eventually settled on the ground due to the effects of wind and rain. During the FDNPP accident, pediatric thyroid screening of children under 15 years of age (approximately 1,080 children) was conducted using NaI(Tl) scintillation survey meters [2, 3].
Several investigative reports have been made on the simple measurements performed after the accident [4–8]. According to those reports, the simple measurement is reported to contain several uncertainties in the measured values, including the timing of measurement, effects of radiation from environmental contributions, effects of radioactive materials adhering to the body surface, and reading errors that occur when the measured values are read. Because once 131I is taken up by the body, it is subject to age-specific biokinetics (biological half-life), which results in its remaining time in the body being shorter than its physical half-life [9, 10], a measurement method that can detect very small amounts of 131I accumulated in the thyroid gland is required.
There are several factors to consider regarding pediatric thyroid screening following a nuclear accident. Easy measurement is required early after the accident in terms of equipment and personnel. Second, the ability to measure thyroid radioactivity is necessary even if the surrounding environment is contaminated. Furthermore, the ability to selectively measure multiple radionuclides is required during an accident. Radiation emitted from the thyroid gland must be detectable at the low doses required for radiological protection. From these perspectives, a NaI(Tl) scintillation survey meter is a simple instrument, but it does not fulfill all these requirements because it does not allow energy analysis.
More than a decade after the FDNPP accident, these screening issues remain unresolved. However, in pediatric thyroid screening, a simplified measurement method is considered the current standard, although it is still affected by the abovementioned uncertainties. In 2023, a manual describing the simplified measurement method was published by the Nuclear Regulatory Commission [11]. While this method has the advantage of simplicity, it does not meet all of the above requirements.
A new measurement device for pediatric thyroid screening has been developed by the Quantum Science Research and Development Organization [12]. This device utilizes a Gd3(Ga,Al)5O12(Ce) (GAGG) scintillator in the detection section and is unique in that it is designed to conform to the shape of the neck. Because the device was developed specifically for pediatric thyroid screening, improved measurement accuracy can be anticipated compared with NaI(Tl) scintillation survey meters. However, it is believed that practical implementation will require time due to the associated costs and training required for deployment of the device in the field.
One type of instrument used for environmental radiation monitoring is the NaI(Tl) spectrometer. Unlike the NaI(Tl) scintillation type survey meter, the NaI(Tl) spectrometer analyzes the energy of each nuclide, enabling measurements focused specifically on 131I. These devices are thus useful in situations involving mixtures of multiple types of radioactive materials, such as nuclear power plant accidents [13]. This study aimed to establish a highly accurate measurement method for pediatric thyroid screening suitable for implementation soon after accidents, considering equipment and personnel constraints. Simplified measurements using NaI(Tl) type survey meters cannot discriminate gamma-ray energies from different nuclides, which complicates efforts to improve accuracy when radionuclides other than 131I affect the measurement. Hence, our study examined the use of NaI(Tl) spectrometers, which can selectively measure multiple radioactive nuclides, and we used the resulting data to calculate the detection limits for screening children aged 1 to 15 years. Monte Carlo simulations were conducted using Particle and Heavy Ion Transport code System (PHITS) version 3.24, which accurately reproduces the behavior of photons, electrons, and positrons over a wide energy range [14].
Monte Carlo simulations have been used frequently since the 1960s to reproduce energy spectra [15–17]. At that time, however, the accuracy of Monte Carlo simulations of the transport processes of low-energy electrons was insufficient to reproduce the energy spectrum [18–21]. Although the accuracy of Monte Carlo simulations has since improved, as has the reproducibility of energy spectra [22–25], simulations do not yet fully agree with experimental data. A previous study by Zerby et al. [15] showed that the level of scintillator emissions, which are responsible for non-linearity in reproducing the energy spectrum, varies with energy. In our study, we investigated the possibility of improving the detection limit and practical performance of the NaI(Tl) spectrometer as an alternative to the NaI(Tl) scintillation survey meter as a device for pediatric thyroid screening after a nuclear power plant accident.
Materials and Methods
1. Simulation of the 131I Energy Spectrum
When reproducing energy spectra using the PHITS, it is preferable to consider factors such as fluorescence efficiency, quantum efficiency, and rates of amplification by dynodes or amplifiers. However, the ability to reproduce internal conditions of a device varies with each device setting and can be complex. If the output of the PHITS is in the format of the number of absorbed doses per unit of energy, the absorbed dose can be treated as an output proportional to fluorescence efficiency, quantum efficiency, and the amplification factor, but this requires a matching proportionality constant. As each nuclide exhibits a different amplification rate, energy spectra are reproduced by matching them with a simple adjustment factor corresponding to the proportionality constant for each nuclide. In addition, NaI(Tl) scintillators may exhibit different indicated values in the case of sudden temperature changes. In this study, experiments were conducted in a constant temperature environment to minimize the effects of temperature on the indicated values.
The first parameter that must be matched is the output of the NaI(Tl) spectrometer (energy spectrum) with the output of the PHITS. Although previous studies have reproduced energy spectra using Monte Carlo simulations, it remains difficult to perfectly match an energy spectrum because its shape can be complicated by the target nuclide, measurement conditions, and other factors [22–25]. The present study utilized sources that were actually available (133Ba, 137Cs, and 40K). Although it is desirable to use an 131I source, many previous studies have used 133Ba as an alternative source because 131I is generally difficult to obtain and its short half-life (8 days) makes its use impractical [26]. A 133Ba source is suitable for verifying energy spectra because the energy of the emitted gamma rays (0.356 MeV) is similar to that of 131I (0.365 MeV).
We measured the energy spectra of the three nuclides using a NaI(Tl) spectrometer and attempted to match the energy spectra with the geometric configuration of the same conditions in the PHITS. An EMF 211 (EMF Japan Co., Ltd.) NaI(Tl) spectrometer (scintillator: length 7.62 cm×diameter 7.62 cm) was used as the measuring device. Energy spectra were calculated for the 133Ba, 137Cs, and 40K sources by varying the source-detector distance from 3 cm to 20 cm. The applied voltage of the spectrometer was fixed at 624 V, which was the manufacturer’s default value. Experimental values were obtained by adjusting the gain such that the total absorption peak of each source was displayed at approximately 700 channels. Figs. 1 and 2 show the experimental layout and detector configuration, respectively. The source was positioned at a height of 50 cm from the installation table to avoid the effect of scattered radiation due to the geometric arrangement. Fig. 3 shows the shape and radioactivity of the sources used in the experiment. For 133Ba and 137Cs, a circular surface source was used, with radioactivity levels of 6.3×105 Bq and 6.9×103 Bq, respectively. Potassium chloride (414 g) sealed in a Marinelli container was used for energy calibration. For 40K, the weight of the nuclide was determined, and the radioactivity was calculated as 6.7×103 Bq using the isotope abundance ratio.
Experimental layout.
Spectrometer probe.
Types and shapes of radiation sources used in the experiment.
The PHITS was originally developed as a program for use in Monte Carlo simulations of radiation behavior. The results of radiation behavior (i.e., fluence and absorbed dose, which change when the radiation interacts) can be modeled. Because the output differs from that of a NaI(Tl) spectrometer, it is necessary to devise a means of matching the output with the energy spectrum. We therefore set up structure in the PHITS version 3.24 that closely resembled the geometric configuration of the detection unit of the NaI(Tl) spectrometer used in the experiment (scintillator: length 7.62 cm×diameter 7.62 cm; aluminum cloak: 2 mm; aluminum casing: 1.5 mm). The radiation source was configured to be of the same size as the actual source and placed at 10,000 randomly selected points within the shape. Simulations were performed for three nuclides (133Ba, 137Cs, and 40K) under the same geometric conditions as in the experiment and repeated for 10 million radiation histories.
The PHITS was configured such that the count rate per absorbed dose, corresponding to the energy absorbed by the scintillator, served as the output. This output merely represents information regarding the physical interaction of the radiation as determined by the geometric arrangement of the source and detector and the energy of the radiation. The output measured by the NaI(Tl) spectrometer is a signal exhibiting variation similar to a Gaussian distribution, which requires conversion from the simulation results. Therefore, we introduced a Gaussian filter in the statistical software R version 4.2.1 (R Foundation for Statistical Computing), which enabled reproduction of a Gaussian distribution by inputting parameters to reproduce the energy spectrum. The parameters included the half-width of the total absorption peak for each energy of radiation and the peak width that varied with the half-width. The full-width at half-maximum (FWHM) was determined using the FWHM values obtained for each radiation source in the experiment (133Ba: 8.3%; 137Cs: 6.6%; 40K: 4.5%). Using a Gaussian filter, it is possible to align the shapes of all absorption peaks. However, this process alone is not sufficient to align the scales of the entire energy spectrum. Because the amplification rate of the NaI(Tl) spectrometer varies depending on the applied voltage and gain settings of the device, energy-dependent fitting parameters are required for the energy spectrum of each source if the gain is adjusted for each nuclide. The adjustment factor used to match the count rate of total absorption peaks (ME) obtained in the experiment with the count rate of total absorption peaks (MP), which is given a Gaussian distribution by a Gaussian filter in the PHITS output, is defined by Equation (1):
To verify the consistency of the results, the experimental data and output of the PHITS to which the Gaussian filter was applied were superimposed and graphed using the statistical software R version 4.2.1. The area of all absorption peaks (sum of the count rates of all absorption peaks) was calculated using Equation (2) to verify the error.
where nex is the sum of the total absorption peak areas obtained in the experiment; np is the sum of the total absorption peak areas simulated by the PHITS.
Although the error can be determined based on the shape of the total absorption peak, in this study, agreement was verified using the ratio of the areas corresponding to the sum of the count rates of all absorption peaks, referring to previous studies that showed agreement using the peak areas [22].
2. Calculation of the NaI(Tl) Spectrometer Detection Limit
In this section, we describe the method used to calculate the detection limit. The effect of ambient dose rate on the detection limit was investigated by calculating the sum of background count rates corresponding to the total absorption peak width of 131I. Using a 137Cs source, the ambient dose was reproduced to vary the dose rate from 0.06 μSv·hr−1 to 1.0 μSv·hr−1. An energy spectrum corresponding to each dose rate was then obtained and used to calculate the detection limit.
The detection limit was determined using the formula shown in Equation (3), focusing on the energy spectrum of environmental contributions equivalent to the channel width of the 131I energy spectrum. The sum of the background count rates, corresponding to the portion equivalent to the peak width of the target nuclide, was used to calculate the standard deviation (σ), and values above 3σ were treated as the detection limit (counts per second [cps]) [27]:
where t is counting time (second); B is the sum of background count in the area corresponding to the peak width of the nuclide of interest. The same counting time for both the ‘sample’ and the ‘background’ must be used.
3. Calculation of the Number of Detectable Days for 131I Accumulated in the Thyroid Gland
Simulations were then conducted using the mathematical phantom shown in Table 1 and Fig. 4, tailored to the ages of actual evacuees (1, 5, 10, and 15 years old), for 131I intake up to approximately 2 months after the accident. The dimension of the mathematical phantom is listed in Table 1 and Fig 4. The mathematical phantoms used here were generated with reference to the Medical Internal Radiation Dose (MIRD) phantom [28]. Size and depth are the most important factors that must be accounted for in Monte Carlo simulations of 131I accumulated in the thyroid gland. The thyroid gland is shaped like a butterfly with spread wings and composed of right and left lobes and an isthmus. In terms of size, 1-year-old children are small and deep, growing larger and closer to the surface as they grow older. The size of the thyroid at various ages is as follows: length to width (1 year old, 1.0–1.5 cm; 5 years old, 1.5–2.0 cm; 10 years old, 2.0–3.0 cm; 15 years old, 3.0–3.0 cm). The larger the size of the thyroid gland, the lower the influence of radiation compared with a smaller gland due to the geometric efficiency of the radiation entering the detector. For this reason, the simulation was performed using a slightly wider calculation range, taking computational efficiency into consideration. With regard to depth, as the thyroid gland is about 1 cm thick, and it is more difficult to detect 131I if it accumulates on the back of the thyroid gland, we assumed that 131I accumulates on the back of the thyroid gland and set the depth to 3.0 cm from the surface of the neck in both cases [26, 29].
Size of Each Phantom
In pediatric thyroid screening using an NaI(Tl) spectrometer, the large probe diameter makes it difficult to adhere the probe to the neck of younger children for measurement. For 1-year-old children, measurements are assumed to be taken 2 cm away from the neck. For children aged 5 years and above, measurements are assumed to be taken with the detector in close contact with the neck. When reproducing the energy spectrum of 131I, an adjustment factor for distance is required. Since the depth of the thyroid gland is assumed to be 3 cm, an adjustment factor of 5 cm was used to reproduce the energy spectrum for 1-year-old children, and an adjustment factor of 3 cm was used for 5- to 15-year-old children. Although a longer measurement time would have resulted in more-accurate measurements, the measurement time was set at 3 minutes, taking into account the physical strength of the test takers, as children in the targeted age groups are young and cannot sustain concentration for long periods, and the probe is large in size and heavy. The energy resolution for 131I (E: 0.365 MeV) could not be measured because no 131I source was available. Therefore, because the energy of 133Ba (0.356 MeV) is similar to that of 131I (0.365 MeV), the energy resolution used for the 133Ba source was applied and calculated. Uptake of 131I into the thyroid gland and residual thyroid radioactivity were calculated according to the method used in our previous study [8]. The equivalent thyroid dose considered to increase the risk of thyroid cancer is 100 mSv [30], and the intake of 131I and residual thyroid radioactivity vary with age. The internal dose to the thyroid was calculated using the dose coefficients described in International Commission on Radiological Protection (ICRP) Publication 71 [31], which is recommended by the ICRP, and residual thyroid radioactivity was calculated using Monitoring to Dose Calculation (MONDAL), published by the Quantum Science Research and Development Organization for calculating the residual thyroid rate for various ages [32]. The obtained residual thyroid radioactivity values were used to calculate the peak count rate of 131I when measured using an NaI(Tl) spectrometer for the period from the intake date to approximately 2 months later using a Monte Carlo simulation. Detection limits were compared with the simulation results to calculate the number of detectable days.
4. Verification of the Effects of 132Te, a Short-Lived Radionuclide
In early monitoring after a nuclear accident, short-lived nuclides released into the environment may affect measured values. The types and ratios of radionuclides released during the FDNPP accident were reportedly determined from analyses of radioactive contamination on the clothing of evacuees [4–6]. Among the detected radionuclides, the total absorption peak of 132Te is similar to the energy of 131I, which could potentially lead to interference with accurate counting of the total absorption peak of 131I. Therefore, using the PHITS, we reproduced the energy spectra of eight radionuclides on the clothing of evacuees during the actual FDNPP accident and verified whether the total absorption peaks of 131I and 132Te could be separated and measured.
Results
1. Verification of Agreement between Experimental Results and Monte Carlo Simulations
Energy spectra both experimentally measured and calculated using the PHITS were superimposed using the statistical software R version 4.2.1, as shown in Figs. 5–7. Simulations were performed for the three nuclides under the same conditions as used in the experiment, varying the measurement distance from 3 cm to 20 cm. The shapes of all absorption peaks were in agreement, within an error of 0.1%–0.7%, except for 133Ba. For 133Ba, the sum effect appeared at closer distances, with errors ranging from 0.5% to 44.0%. Fig. 7E and 7F show the results of a simple correction for the effect of the 133Ba thumb effect using the area ratio. Adjustment coefficients were calculated when correcting for the sum effect: 0.79 and 0.86 for 3 cm and 5 cm, respectively.
Comparison of experimental and simulated 137Cs energy spectra for different measurement distance of (A) 3 cm, (B) 5 cm, (C) 10 cm, and (D) 20 cm.
Comparison of experimental and simulated 40K energy spectra for different measurement distance of (A) 3 cm, (B) 5 cm, (C) 10 cm, and (D) 20 cm.
Comparison of experimental and simulated 133Ba energy spectra for different measurement distance of (A) 3 cm, (B) 5 cm, (C) 10 cm, (D) 20 cm, (E) 3 cm with correction, and (F) 5 cm with correction.
2. Energy Spectrum of 131I Reproduced Using the PHITS
For 1-year-old children, an adjustment factor of 0.86 was used at thyroid depths of 3 cm and 2 cm from the neck surface, whereas an adjustment factor of 0.79 was used at a thyroid depth of 3 cm for 5- to 15-year-old children in order to reproduce the 131I energy spectrum. Fig. 8 shows the results of the reproduced 131I energy spectrum. Because the channel width at which the total absorption peak displayed was equal even if the adjustment factor differed, the results for 1-year-old children are presented here. The total absorption peak at 0.365 MeV emitted from 131I is shown at 699–864 channel, and this peak was clearly distinguishable from the other peaks.
Energy spectrum of 131I calculated by Particle and Heavy Ion Transport code System (PHITS).
3. Experimentally Determined Detection Limits of the NaI(Tl) Spectrometer
Fig. 9 shows the energy spectrum of the environmental contributions varying from 0.06 μSv·hr−1 to 1.0 μSv·hr−1 at a measurement time of 3 minutes determined using the 137Cs source. The detection limit was defined as three times the standard deviation calculated from the sum of the count rates of the energy spectra of the environmental contribution in the range 699–864 channel, which was the same as the total absorption peak of 131I shown in blue in the figure. The values were 0.94 cps for an environmental contribution of 0.06 μSv·hr−1, 1.39 cps for 0.2 μSv·hr−1, 1.65 cps for 0.4 μSv·hr−1, and 2.14 cps for 1.0 μSv·hr−1.
Energy spectra of different environmental contribution: (A) 1.0 μSv·hr−1, (B) 0.4 μSv·hr−1, (C) 0.2 μSv·hr−1, and (D) 0.06 μSv·hr−1.
4. Calculation of Thyroid Uptake, Residual Thyroid Radioactivity, and the Number of Measurable Days
Uptake into the thyroid following inhalation of 131I equivalent to 100 mSv of equivalent thyroid dose for each age group was calculated with reference to the dose coefficients recommended in ICRP Publication 78 [31]. The values were 7,000 Bq for 1-year-old children, 11,800 Bq for 5-year-old children, 24,100 Bq for 10-year-old children, and 36,900 Bq for 15-year-old children. Thyroid residuals were determined using MONDAL, a database published by the Quantum Science Research Organization [31].
Fig. 10 shows the results of a PHITS simulation of evacuees (1, 5, 10, and 15 years old) who inhaled 131I equivalent to 100 mSv of thyroid equivalent dose during evacuation. Comparing the detection limits for each age and considering the number of measurable days and corresponding residual thyroid radioactivity [Bq], under the condition of an environmental contribution of 0.2 μSv·hr−1, for a 1-year-old child, measurements could be taken for up to 38 days (residual thyroid radioactivity: 85 Bq) after inhalation of 131I when measured 2 cm away from the neck. Furthermore, for 5-, 10-, and 15-year-old children, measurements taken with the detector in close contact with the neck allowed detection for up to 50 days (residual thyroid radioactivity: 61 Bq), 67 days (residual thyroid radioactivity: 47 Bq), and 72 days (residual thyroid radioactivity: 48 Bq), respectively, after inhalation of 131I.
The number of measurable days and residual thyroid radioactivity in (A) 1-year-old, (B) 5-year-old, (C) 10-year-old, and (D) 15-year-old children.
5. Effect of the Short-Lived Nuclide 132Te on the Energy Spectrum
During the FDNPP accident, several different radioactive materials were released. Early pediatric thyroid screening is likely affected by short-lived radionuclides. Fig. 11 depicts the results of simulations conducted using a NaI(Tl) spectrometer on the first day after inhalation of 131I, the nuclide that exhibits the most significant impact due to its short half-life. The full absorption peaks of both 132Te and 131I were clearly distinguishable and distinctly displayed.
Energy spectra of eight nuclides (nuclides contaminated on the body surface).
Discussion
1. Agreement between Experimental Values and Monte Carlo Simulations
The experimental values of 137Cs at distances ranging from 3 cm to 20 cm were compared with those derived from Monte Carlo simulations. Fig. 5 demonstrates that the calculated results fit the shape and area of the entire absorption peak, with errors within the range of 0.1% to 0.7%. In the case of 133Ba, this nuclide emits gamma rays at multiple energy levels, which results in a significantly more complex energy spectrum. The error in the area of the total absorption peak ranged from 0.5% to 44.0%, with the error increasing with distance. The error is thought to be attributable to the sum effect being particularly pronounced when multiple gamma rays are emitted, as was the case here. This sum effect became more pronounced when the counting rate was high and the distance was short. The 133Ba source used in this study exhibits high-level radioactivity. When compared with results for the same condition at varying distances, a sum peak clearly not emitted was observed in the portion of the energy. This was due to the high emission rate of characteristic X-rays in the low-energy region, probably influenced by the sum effect with these characteristic X-rays. The present study reproduced the original 133Ba energy spectrum by determining the difference between the area of the total absorption peak obtained experimentally at a distance of 3 cm with the area of the total absorption peak obtained from PHITS simulations, correcting for the difference. Comparing the energy spectrum of 133Ba obtained using the PHITS with the experimentally obtained total absorption peak of 133Ba at 0.356 MeV, the area error at a distance of 3 cm and 5 cm was 44% and 21%, respectively. Therefore, a simple correction for the sum effect was made by multiplying the output of the PHITS by a factor of 1.77 and 1.26 at a distance of 3 cm and 5 cm, respectively. A previous study reported an error of 2.0% in the validation of peak area agreement, with similar results obtained at distances of 10 cm and 20 cm [24]. Therefore, it was possible to simulate the energy spectrum of 131I after successfully reproducing the energy spectra of the three nuclides examined in this study.
2. Reproduction of the 131I Energy Spectrum
The channel width of the total absorption peak for the target nuclide is an essential value for calculating the detection limit. It is important to exactly reproduce the position of the total absorption peak of 131I in the energy spectrum. We therefore used adjustment factors for each energy and distance to reproduce the energy spectrum of 131I. Because the energy of 133Ba (0.356 MeV) is similar to that of 131I (0.365 MeV), adjustment coefficients for 133Ba were used. For 1-year-old children, measurements were taken 2 cm away from the neck, and for children aged 5 years and above, measurements were taken with the detector in close contact with the neck, using values of 5 cm and 3 cm, respectively. As a result, we were able to reproduce the energy spectrum of 131I and confirmed that the channel width of the total absorption peak appeared at positions ranging from 699 ch (0.350 MeV) to 864 ch (0.432 MeV).
3. Detection Limit of a NaI(Tl) Spectrometer
It is useful to calculate the detection limit of the measurement device actually used for screening. We calculated the energy spectrum when varying environmental contributions from 0.06 μSv·hr−1 to 1.0 μSv·hr−1 to determine the detection limit. If the total count rate at the position corresponding to the channel width of the total absorption peak of 131I exceeded three times the standard deviation calculated from the sum of the count rates attributed to environmental contributions, this value was considered the dose from the thyroid and treated as the detection limit [27].
The calculated detection limits were 0.94 cps for an environmental contribution of 0.06 μSv·hr−1, 1.39 cps for 0.2 μSv·hr−1, 1.65 cps for 0.4 μSv·hr−1, and ≥2.14 cps for 1.0 μSv·hr−1, indicating detectability when the count rate exceeds these values. The reason for such high measurement accuracy is that the NaI(Tl) spectrometer focuses solely on the channel width of the total absorption peak of 131I. The effect of environmental radiation can be minimized by limiting the measurement to the count rate from the Compton scattering within the channel width of interest.
In the case of NaI(Tl) scintillation survey meters that detect radiation from all nuclides, an increase in the count rate leads to larger fluctuations. However, the NaI(Tl) spectrometer exhibits a low degree of fluctuations in count rate and provides highly accurate measurements. In the case of a nuclear accident in which multiple radionuclides are released into the environment, use of a NaI(Tl) spectrometer is thus preferable.
4. Calculation of Days Available for Measurement of Residual Thyroid Radioactivity
It is useful to calculate the number of days 131I thyroid radioactivity can be detected based on the detection limit. We simulated acute intake of 131I equivalent to 100 mSv and compared this condition with the detection limit up to 2 months after exposure based on thyroid uptake and residual radioactivity. Comparing with the detection limit, measurements were possible up to 38 days post-inhalation (thyroid retention activity: 85 Bq) for children aged 1 year and 50 days (thyroid retention activity: 61 Bq), 67 days (thyroid retention activity: 47 Bq), and 72 days (thyroid retention activity: 48 Bq) after inhalation for children aged 5, 10, and 15 years, respectively. It was found that at a younger age and with low thyroid intake, the number of measurable days decreased. Therefore, it is desirable to measure residual radioactivity early after inhalation. However, it should be noted that it is difficult to conduct measurements on the day of inhalation because the radioactivity has not yet accumulated in the thyroid gland; the decay curve of residual thyroid radioactivity differs because the uptake and excretion rates vary with age.
5. Effect of the Short-Lived Nuclide 132Te on Measurements
During a nuclear power plant accident, various types of radioactive substances are released. Most of the energy emitted by radioactive substances is >0.5 MeV, which affects the Compton region but has little impact on the overlap with the full absorption peak. However, 132Te emits gamma rays with an energy of 0.228 MeV, similar to the 0.365 MeV energy of gamma rays emitted by 131I. Therefore, we investigated the extent of overlap between the total absorption peaks of 131I and 132Te when these nuclides were mixed. The simulated energy spectra were derived from the ratios of eight radionuclides detected on the clothing of evacuees following the FDNPP accident.
Given the minimal overlap of total absorption peaks for each nuclide, thus allowing for separate observations, it was postulated that 131I could be selectively detected even when measuring early after inhalation. In the case of the FDNPP accident, simplified measurements were recommended for environments exhibiting radiation contributions of <0.2 μSv·hr−1 [2, 3, 11]. However, few areas exhibiting contributions <0.2 μSv·hr−1 were identified. Early measurements conducted after an accident cannot avoid the effect of short-half-life nuclides such as 132Te. The results of our study demonstrated that early measurement allows for the detection of 131I even in the case of an environmental contribution of 1.0 μSv·hr−1. Our data indicate that the use of NaI(Tl) spectrometers enables the detection of thyroid radioactivity following the acute inhalation of 131I equivalent to 100 mSv, even in situations exhibiting environmental contributions ≥0.2 μSv·hr−1.
6. Comparison of NaI(Tl) Scintillation Survey Meters and NaI(Tl) Spectrometers
Finally, the degree to which the accuracy of measurements could be improved by using a NaI(Tl) spectrometer over a NaI(Tl) scintillation survey meter was examined. Table 2 summarizes the characteristics of each of the NaI(Tl) type survey meter and NaI(Tl) spectrometer. Previous studies clearly showed that uncertainty in measured values is unavoidable in cases of simple measurements using NaI(Tl) scintillation survey meters. Furthermore, due to the lack of energy analysis, it is difficult to selectively identify 131I when radioactive materials other than 131I adhere to the body surface. For 1-year-old children, the lower limit of measurable residual thyroid radioactivity is about 600–700 Bq, and 131I can be measured up to 21 days after inhalation [8]. When using the NaI(Tl) spectrometer, 131I can be judged to have been detected if the count rate is greater than or equal to the detection limit of 1.39 cps calculated using Equation (3). For a 1-year-old child, the depth of the thyroid gland is assumed to be 3 cm below the body surface and the measurement position is 2 cm away from the neck surface, in which case the lower limit of possible residual thyroid radioactivity that can be measured is 85 Bq. For a 5-year-old child, the depth of the thyroid gland is assumed to be 3 cm below the surface of the neck, and the measurement position is assumed to be in close contact with the neck. In this case, the lower limit of the residual thyroid radioactivity that can be measured is 61 Bq. These data indicate that the NaI(Tl) spectrometer has a longer detectable period than the NaI(Tl) scintillation survey meter, with the NaI(Tl) spectrometer extending the detectable period up to 38 days after inhalation. However, the total counting time for sample and background is assumed to be 2 minutes for the NaI(Tl) scintillation survey meter and 3 minutes for the NaI(Tl) spectrometer. Although a longer measurement time improves measurement accuracy, we believe that the above measurement time is appropriate considering the burden on the measurer due to the weight of the measurement device.
Comparison of the NaI(Tl) Scintillation Detector and NaI(Tl) Scintillation Survey Meter
7. Limitations of the Study and Future Prospects
In the case of measurement environments characterized by high counting rates or cases involving measurement of highly radioactive sources, sum effects and pile-up cannot be ignored when using a NaI(Tl) spectrometer. This study did not take into account thumb effects or pile-up. In environments in which radiation levels were measured at the time of the FDNPP accident, the sum effect was considered to have negligible impact and did not significantly affect the results of this study. However, caution should be taken if it is not possible to avoid conducting measurements in environments with high ambient dose rates.
Conclusion
Our study revealed that the energy spectrum of NaI(Tl) spectrometers can be reproduced with good accuracy using Monte Carlo simulation and that the use of NaI(Tl) spectrometers for pediatric thyroid screening enables the separation of short-half-life nuclides even with early measurements. NaI(Tl) spectrometers also can provide greater measurement accuracy compared with NaI(Tl) scintillation survey meters. In this study, we found that 131I can be measured up to 38 days after inhalation when measured under the environmental contribution of 0.2 μSv·hr−1, and that the detection limit of the NaI(Tl) spectrometer is 1.39 cps under the same conditions.
Notes
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Ethical Statement
This article does not contain any studies with human participants or animals performed by any of the authors.
Author Contribution
Conceptualization: Kai M. Methodology: Kitajima T, Kai M. Data curation: Kitajima T. Formal analysis: Kitajima T. Supervision: Ojima M, Shinagawa Y, Kai M. Visualization: Kitajima T. Software: Kitajima T. Writing - original draft: Kitajima T. Writing - review & editing: Ojima M, Shinagawa Y, Kai M. Approval of final manuscript: all authors.