TY - JOUR
AU - Kormoll,, Thomas
AB - Abstract The concept of an active dosimetry system for pulsed radiation dose rate measurements is presented. Real-time distinction of pulsed and non-pulsed radiation contributions is based on the time structure of a single interaction. A fast tissue equivalent plastic scintillator is exploited to minimize the pile-up effect influence on absorbed energy measurements. Being connected to a fully digital signal processing board, the detector creates an active dosimetry system with adjustable parameters. With this system, absorbed dose rate measurements were carried out in a photon field with a time structure mimicking a radiotherapeutic beam, but also in the presence of a constant radiation field. Measurements show a linear dependence of a pulsed radiation contribution on the accelerator current in the investigated range of the total dose rate up to 8 μGy h−1. While increasing the accelerator current by 1 μA, the pulsed radiation dose rate grows by (26.2 ± 0.9) nGy h−1 when considering pile-up events. INTRODUCTION In many practical applications (e.g. radiation beam therapy, X-ray diagnostics, research accelerators, laser-induced X-rays, etc.), the radiation field is pulsed. In all facilities where radiation is generated or used, the whole process has to be accompanied with the appropriate dosimetry. Speaking about area monitoring, one needs an active dosimetry system that provides direct readings of the ambient dose equivalent(1), and can immediately warn users in case of exceeding established dose or dose rate limits. Passive dosemeters do not meet these requirements due to time-consuming measurements and, in some cases, a lot of parameters to be involved(2). For continuous radiation, one can exploit a number of conventional well-tested active dosemeters based on ionization chambers, proportional counters or semiconductor diodes. But several studies(3–6) have shown that whereas counting detectors suffer from the dead-time behavior, ionization chambers can be affected from recombination loss in high-intensity fields, while they are of limited suitability in low-intensity field due to their low signal level. The intrinsic energy dependence of semiconductor diodes makes their usage in scattered radiation or non-calibrated beams measurements quite challenging. Thus, an ideal detector for pulsed radiation dosimetry shall overcome these disadvantages and also has to provide measurements of operational dose quantities for the entire range of radiation energy. This paper will give the main considerations regarding the concept of such a detector and its operational regime. The way of distinguishing pulsed and non-pulsed dose contributions in real-time dosimetry will be discussed, and the results of dose rate measurements in pulsed radiation fields imitating radiation fields outside the bunker of a medical LINAC will be presented. MATERIALS AND METHODS Detector To deal with the energy dependency that can also be related to the pile-up and intrinsic dead-time behavior, a tissue equivalent material for a detector is needed. Then, in the energy range where the tissue equivalence condition is fulfilled, simultaneous energy depositions in the detector material from multiple photons (pile-up) will lead to the proportional value of the energy deposition as from summing up contributions from separately coming photons for a soft tissue. After performing an energy calibration and applying a corresponding quality factor (Q = 1 in the case of photon radiation) to the total energy deposition in a tissue equivalent material, one can estimate detector’s equivalent dose which, in the same exposure conditions, will be linearly proportional to the equivalent dose for a soft tissue. This means that the pile-up is not of the concern if a tissue equivalent detector is used as higher weighting of pile-up events is achieved through their higher energy values. Similar reasoning is not valid for count rate evaluations where the knowledge of the pile-up multiplicity and the dead-time behavior are essential for the introduction of appropriate corrections(7–9). The second point is that the dosemeter has to be fast enough to provide time-resolved dosimetry necessary, for example, in radiotherapy. This requires a detector able to measure in the timescale of microseconds. In this regard, fast scintillators, in principle, are not limited if appropriate hardware is used(4,10). Considering the abovementioned, solid plastic scintillators can be chosen as detector materials. Besides having a response time in the range of nanoseconds, they are not affected by pressure and humidity fluctuations and can be easily made in a desired shape and size. To investigate their tissue equivalence, the comparison of mass energy-absorption coefficients of vinyltoluene and polystyrene (polymer bases of many plastic scintillators) and the ICRU four-component soft tissue(1) was made, and the results are shown in Figure 1. One can see that both materials behave in the same way, and differences between them are insignificant. In the region from 200 up to 4 MeV, the ratio of the coefficients can be taken as a constant (0.981 ± 0.004 for vinyltoluene and 0.975 ± 0.004 for polystyrene) which means that simple correction can be done to relate the energy deposition in a plastic scintillator and a soft tissue. Beyond this range, plastic scintillators will underestimate the absorbed dose up to 2.5 times depending on energy, compared with a soft tissue. This pattern is completely satisfactory as behind the treatment room; the overwhelming part of the photon energy spectrum is in the energy region higher than 100 keV(12,13). Figure 1 Open in new tabDownload slide Mass energy-absorption coefficients of the main polymer bases of plastic scintillators and the ICRU four-component soft tissue(11). Figure 1 Open in new tabDownload slide Mass energy-absorption coefficients of the main polymer bases of plastic scintillators and the ICRU four-component soft tissue(11). In this work, the detector (SCIONIX Holland B.V., Regulierenring 5, 3981 LA Bunnik, The Netherlands) consists of the cylindrical (height of 1′′, Ø = 1′′) plastic scintillator EJ-200 (Eljien Technology, 1300 W. Broadway, Sweetwater, Texas 79556, United States) in an aluminum housing of a 0.4-mm thickness, and a Hamamatsu R1924A photomultiplier with a built-in high voltage power supply. Data acquisition and processing The output signal from the detector is sampled and analyzed with a data acquisition board DAQ125 (Serious Dynamics, 01139 Dresden, Germany). The early digitization of voltage pulses allows real-time processing in an FPGA. An event timestamp, an integral over a whole pulse and a pulse shape are determined and transferred into a listmode file. Here, the pulse shape is defined as the ratio of an integral over a part of a pulse to an integral over a whole pulse, i.e. $$\begin{equation} \mathrm{pulse}\ \mathrm{shape}\!=\!\underset{\mathrm{short}\ \mathrm{gate}}{\int}\mathrm{signal}(t)\ \mathrm{d}t/\underset{\mathrm{long}\ \mathrm{gate}}{\int}\mathrm{signal}(t)\ \mathrm{d}t \end{equation}$$(1) that is saved as an eight-bit number where 0 corresponds to 0 and 255 is equal to 0.9961(14). Both short and long gates are of configurable lengths. In this work, they were 20 and 60 samples, respectively. If a pile-up happens (as shown in Figure 2), it leads to the rise of a resultant signal value and increases an integral proportionally to the accounted part of an overlapped event. This is interpreted as the increase of the total energy deposition in the detector. The main problem caused by the pile-up is that when it occurs close to the end of a long integration gate, a major part of an integral is not taken into account in the total energy deposition, which results in considerable energy information losses. Figure 2 Open in new tabDownload slide Integration gate lengths definition and pile-up effect illustration. Two events (pulses from which are shown with dotted and dashed lines) happen within the same integration gate. Integral over the total pulse (solid line) will be saved. The hatched area represents the lost part of the integral over the second pulse and, as a result, unaccounted energy deposition. Figure 2 Open in new tabDownload slide Integration gate lengths definition and pile-up effect illustration. Two events (pulses from which are shown with dotted and dashed lines) happen within the same integration gate. Integral over the total pulse (solid line) will be saved. The hatched area represents the lost part of the integral over the second pulse and, as a result, unaccounted energy deposition. Analyzing timestamps from a listmode file, one can also estimate the probability p of the data loss during the readout. If the signal crosses a threshold, a leading edge trigger fires, and this is considered as a coarse timestamp of an event. After the time equivalent to 224 samples, the rollover of a timestamp counter occurs, and this is saved as a coarse timestamp of 0. The number of rollovers is also tracked by a counter in an FPGA and can be used to compare the number of received rollovers(14), which results in $$\begin{equation} p={n}_l/\left({n}_l+{n}_r\right) \end{equation}$$(2) where nl is the number of lost rollovers and nr is the number of readout rollovers. The rollover is lost when the coarse timestamp of the ith event is smaller than the coarse timestamp of the (i−1)th event, and none of them is equal to 0. Within the integration gate no other trigger can occur, and the shortest time when the trigger is ready again is one sample after the last sample of the long gate. This means completely non-paralyzable behavior with a known dead-time (the long gate length plus one sample(14)) that can be used for dead-time corrections. The revealed three main causes of signal losses during the data acquisition, which are (1) pile-up, (2) readout and (3) dead-time, were in one extent or another considered and explored in the work. Experimental setup Measurements in a pulsed radiation field were carried out at the γELBE (HZDR, Germany). The electron LINAC was operated at the energy of electrons of 18 MeV. A niobium target was used for producing bremsstrahlung. To mimic radiation protection measurements outside the treatment room, the plastic scintillator was placed in a scattered radiation field that keeps the time structure of the initial beam but has an average energy of 100 keV. For this purpose, a PMMA phantom (10 cm × 10 cm × 20 cm, the density of 1.18 g cm−3) was used as shown in Figure 3. Figure 3 Open in new tabDownload slide Principal schema of the experimental setup. The dimensions of the PMMA phantom are indicated. The distance between the 22Na source and the detector was 29 cm. Figure 3 Open in new tabDownload slide Principal schema of the experimental setup. The dimensions of the PMMA phantom are indicated. The distance between the 22Na source and the detector was 29 cm. With such experimental geometry, the field comprises photons scattered at least once of at least 70°. From angular Klein–Nishina distribution of scattered photons, one can deduce that only the low-energy part of the incident beam will reach the detector, while high-energy photons will be mostly scattered at lower angles. Even in the extreme case, when the photon energy tends to the infinity, the energy of such a photon scattered at 70° will tend to 777 keV. Moreover, due to the pair production, one expects an increased amount of 511 keV quanta coming from the positron annihilation. Considering that not only single scattering is possible, the energy of photons hitting the detector will be even smaller. When there are photons of such high energy, a question of photonuclear reactions (especially when generating neutrons) arises. Because of the bremsstrahlung spectrum shape, the intensity of photons with the energy above the neutron threshold on 13N or 15O is almost negligible. Taking into account the abovementioned and the absence of further material in the beam, neutron contamination has not been considered. To imitate the time structure of the radiotherapy beam, the accelerator was working in the macropulse mode when micropulses with widths Δtμ in the range of picoseconds and the frequency fμ of 13 MHz were additionally modulated to form macropulses with the duration Δtm of 5 μs and the period T of 5 ms. The DAQ125 board was synchronized to the accelerator frequency and was running with 13 × 8 MSamples s-1. Measuring of the radiation coming only from the accelerator is rather straightforward. That is why an additional source of continuous radiation was used. For all measurements, a 22Na source was attached to the PMMA phantom opposite to the detector. The main idea was to implement and test an algorithm that would allow real-time distinction of pulsed and non-pulsed radiation contributions to the dose rate. Dosimetry by means of the plastic scintillator Energy calibration of the detector was performed using the wide-angle Compton-coincidence technique(15). This method was chosen because of being a reliable direct method of low-Z materials energy calibration with easy implementation. Plastic scintillator and CeBr3 scintillator acted as scatter and absorber detectors, respectively, and a 137Cs source was used for measurements. The results of energy calibration were verified by comparison of observed Compton edges of such sources as 22Na, 137Cs, 60Co and 207Bi with theoretical values. After the energy calibration, the measured absorbed dose rate |${\overset{.}{D}}_{\mathrm{m}}$| of photon radiation in the plastic scintillator can be calculated as $$\begin{equation} {\overset{.}{D}}_{\mathrm{m}}=\frac{1}{\left(1-p\right) mt}\sum \limits_n\left({aI}_n+b\right) \end{equation}$$(3) where a and b are energy calibration parameters, In is the integral over the whole nth pulse, m is the mass of the sensitive volume of the detector, p is the probability of the data loss during the readout and t is the duration of a measurement. In this so-called multiple pulse mode, the detector accumulates the dose from a sequence of pulses over a period t. As one is not interested in the total number of counts or the count rate to apply (3), no dead-time corrections are needed. Due to the tissue equivalence of the detector material, the pile-up is also not of the concern if all pile-up events completely fall into a long integration gate. As discussed before, to get the absorbed dose in soft tissue, one has to apply a correction factor r. This will result in an absorbed dose rate |${\overset{.}{D}}$| $$\begin{equation} \overset{.}{D}={\overset{.}{D}}_{\mathrm{m}}/r \end{equation}$$(4) where r = (0.981 ± 0.004). Here and after, speaking about dose rate values, one means a corrected absorbed dose rate |${\overset{.}{D}}$|⁠. Before the actual experiment, a test measurement using a 137Cs source with a well-known air kerma rate was performed in the laboratory of the Institute for Nuclear and Particle Physics of the TU Dresden (Germany). The corrected absorbed dose rate was estimated to be (14.0 ± 0.1) μGy h−1. Applying the ratio of mass energy-absorption coefficients for the ICRU four-component soft tissue and air(11), one can calculate that the absorbed dose in air is (13.7 ± 0.1) μGy h−1 which can be equaled to the air kerma as the photon energy is not high. For this source, the stated air kerma rate at the same distance was (15 ± 2) μGy h−1, which shows quite good agreement between these two values at the given energies of the source. As the evaluation was made after the measurement, this resulted in a rather low uncertainty value. RESULTS AND DISCUSSION Discrimination of pile-up events First, the measurements with the 22Na source only were carried out. For this case, Figure 4a represents the relation between the registered energy and the corresponding pulse shape. There is a region where, independent of the registered energy, the pulse shape is stable and is enclosed within certain borders. The left border is clearly pronounced and is equal to 63. The right border was estimated to be 125, thus, 95% of all registered events were within these borders and the rest 5% had the pulse shape greater than 125. Figure 4 Open in new tabDownload slide Registered energy versus corresponding pulse shape for the following cases: (a) is for the 22Na source only used; (b) is for the accelerator switched on and operating at the current of 200 μA and the 22Na source used; (c) is for double pile-up simulation results. The relation between the energy and the pulse shape shown in (a) is used to estimate the limits applied to the pulse shape. The number of pile-up events for (b) is about 15% of the total number of events. The branched structure of the pulse shape distribution is observed in (c). Figure 4 Open in new tabDownload slide Registered energy versus corresponding pulse shape for the following cases: (a) is for the 22Na source only used; (b) is for the accelerator switched on and operating at the current of 200 μA and the 22Na source used; (c) is for double pile-up simulation results. The relation between the energy and the pulse shape shown in (a) is used to estimate the limits applied to the pulse shape. The number of pile-up events for (b) is about 15% of the total number of events. The branched structure of the pulse shape distribution is observed in (c). Pulses with pulse shape numbers more than 125 and small values of energy are considered to be not registered completely. The presence of the pile-up of photons from different micropulses should appear in the region with pulse shape numbers less than 63. If photons originate from the same micropulse, the system cannot distinguish them due to extremely small micropulse width (~ps) in comparison with the scintillator pulse width (~ns). In this case, the energy deposition of the photons will be counted as the energy deposition from one photon with the energy equal to the sum of registered photons energies. Such events will contribute to pulse shape numbers from 63 to 125. Figure 4b shows the same relation as Figure 4a, but the accelerator was switched on and the accelerator current of 200 μA was set. It is seen that a number of events appeared in the region where the pile-up effect was supposed to be observed. Also, additional areas in the form of two branches arose besides the region designated in Figure 4a. This effect can be explained based on the micropulse structure of the macrobunch. As short and long integration gates are of fixed lengths, and the time between two micropulses is about 77 ns, it is possible to estimate at what time after starting the integration the pile-up can occur and the maximum number of possible pile-up events. Based on this, a simplified simulation of a double pile-up was made where detector pulses were modeled as rectangles with the length equal to the long integration gate and heights equal to energies of photons. The double pile-up only was considered as the average number of registered events per macropulse was estimated to be less than 3, which means that for investigated exposure conditions, the probability of the pile-up with the multiplicity higher than two is almost negligible. The exponential energy distribution of photons was assumed. The arrival time of the second event was randomly picked from the array of (0 + 77·iμ) ns where 0 ns is the time of the first event registration, and iμ is an integer from 1 to 7. Events with iμ of 1 or 2 occur within the short integration gate. The results of the simulation are presented in Figure 4c. The final picture is formed by the overlap of distributions (I), (II) and (III) coinciding with the increase of pulse shape numbers. The pile-up with photons from two consequent micropulses (iμ = 1) falls only into the area (I). When iμ = 2, such pile-up forms the second area (II). All pile-up events that take place after the short integration gate are within the area (III). In ideal case, the theoretical pulse shape distribution should be a vertical line around 85 which is equal to the division of integration gates lengths. But due to some factors (not ideally rectangular pulse shape with different rising and falling edges, not accounted delay of the pulse, etc.), a shift and broadening of the pulse shape distribution take place. The assumption is that in the experimental data, the lowest branch corresponds to the two simulated areas (II) and (III) merged, while the middle branch is caused by the pile-up with a photon from a subsequent micropulse (distribution (I)). This also explains the different intensities of the two branches in Figure 4b. Summing up, pulse shape distortions with the increase of the accelerator current can be explained by the double pile-up effect only, without concerning the neutron contamination that may cause pulse shape values different from photons’. Thus, the energy calibration of the scintillator for the photon radiation remains in force, and dosimetry considerations (3) and (4) can be used. Distinguishing pulsed and constant radiation Time difference histograms for cases when the accelerator was operating at the current of 25 and 200 μA with the 22Na source used are shown in Figures 5 and 6, respectively. The peaks with the time difference that is a multiple of the macropulse period correspond to the radiation coming from the accelerator, while the background part coming predominantly from the 22Na behaves according to the Poisson process. Figure 5 Open in new tabDownload slide Time difference histogram for the case of the accelerator switched on and the 22Na source used. The accelerator current was 25 μA. The pulse shape limitation was not applied. Figure 5 Open in new tabDownload slide Time difference histogram for the case of the accelerator switched on and the 22Na source used. The accelerator current was 25 μA. The pulse shape limitation was not applied. Figure 6 Open in new tabDownload slide Time difference histogram for the case of the accelerator switched on and the 22Na source used. The accelerator current was 200 μA. The pulse shape limitation was not applied. Due to a rather high accelerator current value, peaks with the time difference larger than 10 ms do not appear. Figure 6 Open in new tabDownload slide Time difference histogram for the case of the accelerator switched on and the 22Na source used. The accelerator current was 200 μA. The pulse shape limitation was not applied. Due to a rather high accelerator current value, peaks with the time difference larger than 10 ms do not appear. Comparing these two histograms, one can see that with the increase of the accelerator current, the probability not to register photons from the one macropulse within its duration decreases, which manifests disappearance of peaks with the time difference larger than 5 ms. Because of being occupied with the accelerator-originating radiation during the macropulse, the time when the detector can mostly detect the radiation coming from the 22Na source is in between macropulses. That is why the exponential part of the time difference histogram in Figure 6 is limited with 5 ms. Based on the time difference histograms and considering the time structure of the beam, one can distinguish accelerator and non-accelerator generated radiation even in real-time measurements. The pulsed radiation from the accelerator has the time difference that belongs to the intervals of $$\begin{equation} \left[ kT-\Delta{t}_{\mathrm{m}};\kern0.5em kT+\Delta{t}_{\mathrm{m}}\right] \end{equation}$$(5) where k is used for positive integers. If registered events come from the same macropulse, the time difference between them belongs to the interval of $$\begin{equation} \left[\mathrm{long}\ \mathrm{gate};\Delta{t}_{\mathrm{m}}\right] \end{equation}$$(6) In this work, the long integration gate of 60 samples length corresponds to ~577 ns, and Figure 7 confirms it. Figure 7 Open in new tabDownload slide Enlarged peak from the time difference histogram corresponding to events registered within the one macropulse for the accelerator current of 200 μA and the 22Na source used. The pulse shape limitation was not applied. The micropulse structure of the macropulse is seen as the number of small peaks spaced by ~77 ns. Figure 7 Open in new tabDownload slide Enlarged peak from the time difference histogram corresponding to events registered within the one macropulse for the accelerator current of 200 μA and the 22Na source used. The pulse shape limitation was not applied. The micropulse structure of the macropulse is seen as the number of small peaks spaced by ~77 ns. Keeping in mind that the radiation from the 22Na source and background can also fall into (5) and (6), the algorithm was used to estimate these contributions in the count rate and the dose rate. For the measurement with the 22Na source only, so-called ‘pulsed’ parts were 0.2 and 3.7% of the total count rate and the total dose rate, respectively. These values are valid only for the before mentioned beam structure. As soon as the macropulse frequency or duration changes, the intervals (5) and (6) change, respectively, and the effects of the continuous source will be compensated by different fractions of ‘pulsed’ components. In this study, the non-pulsed field was so low in intensity that the pile-up considerations were of no importance for the count rate. However, the same principles of the pile-up accounting (i.e. the energy weighting) as in the pulsed component applied to the dose rate. This means that pile-up events contribute too less to the count rate, but a-priory correct to the calculated dose rate in a tissue equivalent detector material. Assuming the constant probability of the radiation detection, and equivalent exposure conditions, these contributions were considered in further calculations. To validate the proposed algorithm, the count rate from the pulsed radiation was evaluated. Theoretically, the pulsed radiation count rate within one macropulse maximally possible to detect is numerically equal to fμ and should be 13 Ms−1. Because the integration gate is longer than the time between micropulses, the maximal measured count rate without any corrections is about 1.9 Ms−1. Due to the pulsed structure of the beam, conventional dead-time corrections for the count rate are not applicable. Special modifications(7–9) developed for pulsed beams require additional knowledge of the radiation field, e.g. macropulse shape, average pile-up multiplicity, etc. Moreover, they consider the constant intensity within the macropulse, while the micropulse structure of the macropulse is observed (see Figure 7). That is why the pulsed radiation count rate is restored without using dead-time corrections but based on the results of pile-up simulations. As for the investigated exposure condition, only a double pile-up is expected; thus, all events with the pulse shape from the pile-up region are counted twice. Raw and corrected results are shown in Figure 8. Figure 8 Open in new tabDownload slide Pulsed radiation count rate versus accelerator current values. Contributions of the 22Na source and background are subtracted. Uncertainties are enclosed within markers borders. Figure 8 Open in new tabDownload slide Pulsed radiation count rate versus accelerator current values. Contributions of the 22Na source and background are subtracted. Uncertainties are enclosed within markers borders. One can see that even for the greatest accelerator current used in the experiment, the pulsed radiation count rate is still quite far from its maximum theoretical saturation value and behaves linearly. Consideration of the double pile-up increases the count rate from 5 up to 20% in comparison with not corrected values for 25 and 200 μA, respectively. Since the conditions are far from saturation, the further investigation of the relation between the count rate linearity and the pile-up multiplicity, as well as limitations of the discussed way of the pile-up accounting to the count rate, is needed. Pulsed dose rate contribution to the total dose rate The results of total dose rate, pulsed and non-pulsed radiation dose rate calculations taking and not taking into account the pulse shape limitation are shown in Figures 9 and 10, respectively. It worth noting that Figure 10 displays an average pulsed radiation dose rate (calculated for the whole measurement time). Concurrently with this, the dose rate per pulse will be 103 times higher. Both figures contain the values of the dose rate calculated when dismissing pile-up events. This quantity is somehow artificial and introduced here only for the ascertaining the pile-up influence on the dose rate values. Figure 9 Open in new tabDownload slide Total (assuming the 22Na source and background) dose rate versus accelerator current. Uncertainties are enclosed within markers borders. Figure 9 Open in new tabDownload slide Total (assuming the 22Na source and background) dose rate versus accelerator current. Uncertainties are enclosed within markers borders. Figure 10 Open in new tabDownload slide Pulsed (p) and non-pulsed (np) radiation dose rates versus accelerator current. Uncertainties are enclosed within markers borders. Figure 10 Open in new tabDownload slide Pulsed (p) and non-pulsed (np) radiation dose rates versus accelerator current. Uncertainties are enclosed within markers borders. One can notice that even simple adding of pile-up events to the cumulative energy deposition increases the dose rate values dramatically. For the accelerator current of 200 μA, the total dose rate is 26% higher and the pulsed radiation dose rate is 54% higher in comparison with respective cases of limited pulse shapes. For the data obtained without the pulse shape limitation, the increase of the accelerator current by 1 μA leads to the increase of the total dose rate and the pulsed radiation dose rate by (25 ± 1) nGy h−1 and (26.2 ± 0.9) nGy h−1, respectively. The linear behavior of the dose rate with accounted pile-up denotes that for the investigated dose range of several μGy when the double pile-up prevails, the length of the long integration gate is sufficient to prevent noticeable losses of signal due to the pile-up. If the accelerator power continues increasing, the average multiplicity of the pile-up will grow, which means that a great part of pile-up pulses will be lost due to limited integration gate. One of the possible decisions is to expand the long integration gate up to the macropulse duration. In this case, the detector will integrate all events within the macropulse, and pile-up events will be accounted completely. Such a case with a pile-up of high multiplicity will contribute to the dose through the high value of the energy deposited. This approach with increased integration long gate is going to be tested during the next experiment. The behavior of the non-pulsed radiation dose rate is not constant and approaches the reference value (measured with the 22Na source only) while increasing the accelerator current. This can be explained by underestimating the pulsed radiation dose rate at low accelerator current values when mostly one event per macropulse is registered, and this event cannot be definitely related to the accelerator or the background and the 22Na radiation. This means that the proposed approaches (5) and (6) will overestimate the pulsed radiation dose rate contribution in cases with very low accelerator current, while in high pulsed dose rate scenarios, the accuracy of the approach should improve. This question also will be investigated in the next experiment. CONCLUSIONS AND OUTLOOK The presented algorithm of distinguishing pulsed and non-pulsed radiation based on the time difference between consecutively registered events can be easily implemented in the software of any active detector. For the investigated dose rate range of several μGy h−1 in scattered photon fields, the double pile-up assumption is still valid for the count rate corrections. This is proved by both simulation results and linear dependency of the count rate versus accelerator current. In scenarios with higher accelerator power, the relation between the pile-up multiplicity and the count rate will be different. Consequently, without further knowledge of the radiation field, no simple corrections of the count rate can be done. While the count rate is correctly recovered using the information from the pulse shape parameter, the dose rate is calculated by taking obviously pile-up events into account. The required higher weighting of those is achieved through their higher energy values. The weighting in terms of dose is correct, as long as the detector can be considered tissue equivalent. This means that the request of knowing the pile-up multiplicity is not of concern in case of the dose rate, since multiple pile-up events have also higher energy values, which reflect the correct weighting for tissue equivalent scintillators. If all pile-up pulses are registered completely, the dose rate can be restored as long as the data acquisition system is not saturated. One challenge that applies to the improvement of the presented detector is to adjust it for the area monitoring. In radiation protection, the operational dose quantities (H*(10) and/or H´(0.07)) are to be measured(1). Even though for photon radiation, the equivalent dose is numerically equal to the absorbed dose; the corrected absorbed dose rate D cannot be directly used instead of operational quantities due to the differences in their definition. For the H*(10) value, this is mainly because of not equivalent geometrical conditions. If the energy spectrum of incident radiation is monoenergetic, one can go from the absorbed dose in plastic to the absorbed dose (or kerma) in air and then employ the conversion coefficients(16) from the air kerma to the H*(10). It is quite easy to relate the indications of the detector to the H*(10) for a number of standard calibration sources. But in the case of mixed radiation field, the deconvolution of spectrum will be required which is not a trivial tasks no full absorption peaks are observed in low-Z organic scintillators. At the present moment, the Monte Carlo simulation of the detector is going to be done for every reference energy from the ICRP Publication 74(16). The results will allow to optimize the geometry of the detector for that to provide an absorbed dose value directly proportional to the H*(10). Otherwise, a correction function should be derived. One more challenge is the dosimetry of the low-penetrating radiation, e.g. laser-induced X-rays, where the quantity H´(0.07) is preferable to be measured. Here, one assumes the energy range of about 10 keV(2), which means that the housing of the detector will absorb a great part of radiation. Also, the behavior of the detector material in this energy region in comparison with a soft tissue (see Figure 1) is hardly correctable in the online regime without knowing the energy spectrum of the incident radiation. The preliminary idea is to change the material and thickness of the detector housing as well as to decrease the thickness of the scintillator. However, this should be verified based on simulation results. ACKNOWLEDGEMENT Parts of this research were carried out at ELBE at the Helmholtz-Zentrum Dresden-Rossendorf e. V., a member of the Helmholtz Association. We would like to thank the experimental facility staff for assistance. References 1. International Commission on Radiation Units and Measurements . Determination of dose equivalents resulting from external radiation sources. ICRU report 39 . J. ICRU os20 ( 2 ), 18 ( 1985 ). doi: 10.1093/jicru/os20.2.Report39 . OpenURL Placeholder Text WorldCat Crossref 2. Behrens , R. , Pullner , B. and Reginatto , M. X-ray emission from materials processing lasers . Radiat. Prot. Dosimetry 183 ( 3 ), 361 – 374 ( 2019 ). doi: 10.1093/rpd/ncy126 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Ankerhold , U. , Hupe , O. and Ambrosi , P. Deficiencies of active electronic radiation protection dosemeters in pulsed fields . Radiat. Prot. Dosimetry 135 ( 3 ), 149 – 153 ( 2009 ). doi: 10.1093/rpd/ncp099 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Gotz , M . Dosimetry of highly pulsed radiation fields. HZDR, Dresden-Rossendorf, Germany . HZDR-90 ( 2018 ). https://www.hzdr.de/publications/Publ-27317 Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 5. EPD TruDose for dosimetry in short pulsed X-ray fields using the example of industrial pulsed X-ray source XR-200 . https://www.thermofisher.com/order/catalog/product/EPDTRUDOSE#/EPDTRUDOSE 6. Friedrich , S. and Hupe , O. Dose measurements in pulsed radiation fields with commercially available measuring components . Radiat. Prot. Dosimetry 168 ( 3 ), 322 – 329 ( 2016 ). doi: 10.1093/rpd/ncv355 Google Scholar PubMed OpenURL Placeholder Text WorldCat 7. Westcott , C. H. and Elwin , O. M. L. A study of expected loss rates in the counting of particles from pulsed sources . Proc. R. Soc. Lon. A. 194 ( 1039 ), 508 – 526 ( 1997 ). doi: 10.1098/rspa.1948.0094 . Google Scholar OpenURL Placeholder Text WorldCat Crossref 8. Cormack , A. M. Dead-time losses with pulsed beams . Nucl. Instrum. Methods. 15 ( 3 ), 268 – 272 ( 1962 ). doi: 10.1016/0029-554X(62)90086-1 . Google Scholar Crossref Search ADS WorldCat 9. Helbig , K. and Henniger , J. Dead-time losses in pulsed radiation fields . Radiat. Meas. 47 ( 5 ), 383 – 388 ( 2012 ). doi: 10.1016/j.radmeas.2012.03.002 . Google Scholar Crossref Search ADS WorldCat 10. Beierholm , A. R. , Andersena , C. E., Lindvold , L. R. and Aznarb , M. C. Investigation of linear accelerator pulse delivery using fast organic scintillator measurements . Radiat. Meas. 45 ( 3–6 ), 668 – 670 ( 2010 ). doi: 10.1016/j.radmeas.2009.11.023 Google Scholar Crossref Search ADS WorldCat 11. Hubbell , J. H. and Seltzer , S. M. Tables of X-ray mass attenuation coefficients and mass energy-absorption coefficients 1 keV to 20 MeV for elements Z=1 to 92 and 48 additional substances of dosimetric interest. NISTIR . ( Gaithersburg, MD, USA ; Tech. Rep. NISTIR-5632 ) ( 1995 ). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 12. Horst , F. , Fehrenbacher , G. and Zink , K. On the neutron radiation field and air activation around a medical electron LINAC . Radiat. Prot. Dosimetry 174 ( 2 ), 147 – 158 ( 2017 ). doi: 10.1093/rpd/ncw120 Google Scholar PubMed OpenURL Placeholder Text WorldCat 13. Keehan , S. , Taylor , M. L., Smith , R. L., Dunn , L., Kron , T. and Franich , R. D. Dose and gamma-ray spectra from neutron-induced radioactivity in medical linear accelerators following high-energy total body irradiation . Radiat. Prot. Dosimetry 172 ( 4 ), 327 – 332 ( 2016 ). doi: 10.1093/rpd/ncv480 . Google Scholar Crossref Search ADS PubMed WorldCat 14. DAQ125 Firmware (datasheet) . ( Serious Dynamics ). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 15. Roemer , K. et al. A technique for measuring the energy resolution of low-Z scintillators. 2009 IEEE Nucl. Sci. Symp. Conference Record (NSS/MIC) , Orlando, FL , 6 – 11 ( 2009 ). doi: 10.1109/NSSMIC.2009.5401909 . 16. ICRP . Conversion coefficients for use in radiological protection against external radiation. ICRP publication 74 . Ann. ICRP 26 ( 3–4 ), 205 ( 1996 ). OpenURL Placeholder Text WorldCat © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - DOSIMETRY WITH THE ABILITY TO DISTINGUISH PULSED AND NON-PULSED DOSE CONTRIBUTIONS
JF - Radiation Protection Dosimetry
DO - 10.1093/rpd/ncaa120
DA - 2020-10-16
UR - https://www.deepdyve.com/lp/oxford-university-press/dosimetry-with-the-ability-to-distinguish-pulsed-and-non-pulsed-dose-c5XjJUnwqh
SP - 437
EP - 445
VL - 190
IS - 4
DP - DeepDyve
ER -