Technical Note: On the use of cylindrical ionization chambers for electron beam reference dosimetry

Transcription

1 Technical Note: On the use of cylindrical ionization chambers for electron beam reference dosimetry Bryan R. Muir a) and Malcolm R. McEwen Measurement Science and Standards, National Research Council of Canada, Ottawa, ON K1A 0R6, Canada (Received 17 May 2017; revised 26 July 2017; accepted for publication 31 August 2017; published 23 October 2017) Purpose: To investigate the use of cylindrical chambers for electron beam dosimetry independent of energy by studying the variability of relative ion chamber perturbation corrections, one of the main concerns for electron beam dosimetry with cylindrical chambers. Methods: Measurements are made with sets of cylindrical and plane-parallel reference-class chambers as a function of depth in water in 8 MeV and 18 MeV electron beams. The ratio of chamber readings for similar chambers is normalized in a high-energy electron beam and can be thought of as relative perturbation corrections. Data are plotted as a function of mean electron energy at depth for a range of depths close to the phantom surface to R 80, the depth at which the ionization falls to 80% of its maximum value. Additional, similar measurements are made in a Virtual Waterâ phantom with cylindrical chambers at the reference depth in a 4 MeV electron beam. Results: The variability of relative ion chamber perturbation corrections for nominally identical cylindrical Farmer-type chambers is found to be less than 0.4%, no worse than plane-parallel chambers with similar specifications. Conclusions: This work discusses several issues related to the use of plane-parallel ion chambers and suggests that reference-class cylindrical chambers may be appropriate for reference dosimetry of all electron beams. This would simplify the reference dosimetry procedure and improve accuracy of beam calibration. Her Majesty the Queen in Right of Canada Reproduced with the permission of the Minister of Science. [ Key words: electron beams, ionization chamber, reference dosimetry 1. INTRODUCTION 1.A. Electron beam reference dosimetry Reference dosimetry protocols 1 3 recommend the use of plane-parallel chambers for calibration of low-energy electron beams. Following the TG-51 protocol, plane-parallel chambers must be used in beams with nominal energy less than 6 MeV (R cm) and their use is preferred in beams with nominal energy less than 10 MeV (R cm). Other codes of practice make similar recommendations, although the mandatory minimum energy limit varies. There are two main reasons for recommending the use of plane-parallel chambers in low-energy electron beams. A well-guarded plane-parallel chamber is thought to minimize effects from in-scattered electrons and therefore the chamber samples the electron fluence incident only through the front window. This assumption was verified experimentally, although with uncertainties on the order of 1%, by previous authors. 4,5 More recently, Zink et al. 6 used Monte Carlo simulations to demonstrate that in-scattering is not eliminated no matter how wide the guard. The front windows of plane-parallel chambers are typically thin and made from materials with similar properties to water, so the influence of the wall is assumed to be negligible. More recent Monte Carlo calculations 7 have shown that this assumption is not valid, resulting in corrections at the reference depth of up to 1.7% for the NACP-02 chamber and even larger effects deeper in water, especially in low-energy electron beams. The use of cylindrical chambers for electron beam dosimetry has been studied for decades. 4,5,8 11 Cylindrical chambers were not recommended by modern dosimetry protocols 1 3 for reference dosimetry measurements in low-energy electron beams because early publications 8 10 demonstrated large fluence corrections (up to 5%). These studies determined fluence corrections for cylindrical chambers by comparing their response with Fricke dosimetry systems or plane-parallel chambers requiring assumptions about wall and fluence corrections as described above. It was thought that if such large corrections were required, small variations in cavity dimensions (e.g., from manufacturing tolerances) would introduce large uncertainties if generic beam quality conversion factors, k Q, were provided. An additional issue related to electron beam calibration with plane-parallel chambers is that Kosunen et al. 12 reported up to 4% variations for these chambers in 60 Co beams, which would result in large uncertainties if generic k Q factors were provided. To avoid these uncertainties the TG-51 protocol recommended that plane-parallel chambers be cross-calibrated against well-behaved cylindrical chambers in highenergy electron beams. More recently, measurement-based publications have shown that k Q factors for plane-parallel chambers are perhaps not as variable as previously thought However, the longterm stability of plane-parallel chambers is not as good as reported for cylindrical reference chambers, which suggests that a cross-calibration procedure is still the preferred 6641 Med. Phys. 44 (12), December /2017/44(12)/6641/6 Her Majesty the Queen in Right of Canada Reproduced with the permission of the Minister of Science. 6641

2 6642 Muir and McEwen: Cylindrical chambers in electron beams 6642 route to use plane-parallel chambers for reference dosimetry measurements. Recent surveys on reference dosimetry practices have shown that clinical physicists in North America are comfortable using the same cylindrical chambers that they use in photon beams for electron beam calibration. 19 Given the additional complication associated with requiring a crosscalibration procedure to use plane-parallel chambers, the issues related to the behavior of plane-parallel chambers and the demonstrated stability of reference-class cylindrical chambers over decades, we ask the question in this work, Can cylindrical reference-class chambers be used for reference dosimetry of all electron beams, regardless of energy?" 1.B. Relative ion chamber perturbation factors From Spencer-Attix cavity theory, the absorbed dose to water, D w, is related to the absorbed dose to the air in an ion chamber, D ch, through D w ¼ L water D P Q D ch (1) q air where ð L D q Þ water air is the Spencer-Attix water-to-air mean restricted mass collision stopping power ratio with cutoff energy D and P Q is the overall ion chamber perturbation correction to account for perturbations to the electron fluence from the introduction of the chamber walls, air cavity and central electrode (if applicable). The absorbed dose to the air in the ion chamber is also related to the chamber reading, M, through D ch ¼ W M (2) e airm air where ð W e Þ air is the average energy lost per unit charge released by electrons slowing completely in air and m air is the mass of the air in the cavity. If one combines Eq. (1) and (2), takes the ratio of chamber readings for two chambers and normalizes, for example, in a high-energy electron beam with quality Q ecal then ð L D q Þ water air, ð W e Þ air and m air cancel and one obtains M Q 2 ¼ P Q Q;1 : (3) M 1 Q ecal P Q;2 Q ecal That is, the normalized ratio of ion chamber readings is representative of relative ion chamber perturbation corrections. This principle is used in this work to investigate the use of cylindrical chambers in low-energy electron beams by studying the variability of these relative perturbation corrections. 2. METHODS 2.A. Depth-ionization measurements In our previous work, we demonstrated that referencequality data can be derived from electron-beam depthionization measurements. 20 We did this by converting data obtained as a function of depth to data as a function of mean electron energy at depth, E z, and then comparing this to results from other investigations that made measurements only at the reference depth (the assumption being that E z is sufficient to characterize different clinical electron beams). For example, one of the data sets we looked at was the normalized ratio of readings for an NE2571 relative to an NACP-02 chamber obtained by Wittkamper et al., 9 which is relevant for this investigation. This was one of the original publications that suggested large fluence corrections for cylindrical chambers and the agreement between our data and that of Wittkamper et al. was very good. The major advantage of our method is that data at a much wider range of electron energies is obtained (using measurements vs depth and analyzing the data as a function of E z rather than only performing measurements at a fixed depth) with no significant loss in accuracy or precision. With the goal of investigating the variability of cylindrical ion chamber perturbation corrections at low electron energies, we investigate these perturbations as a function of E z. We determine E z here following the IPEM reports. 3,21 Measurements are made here in 8 MeV and 18 MeV electron beams from the NRC Elekta Precise linear accelerator using the method we established in our previous publication. 20 A cm 2 clinical applicator is used to define the beam and two ion chambers are mounted on the applicator for linac monitoring. These are outside of the beam incident on the water phantom. Ion chambers are set up along the beam axis using a 10 cm precision mechanical stand-off positioned against the thin front window of a cm 3 water phantom to define the initial depth of 10.2 g/cm 2 (accounting for the water-equivalent thickness of the front window) in a horizontal geometry. Chambers are preirradiated to 1000 MU before stepping the chamber through the phantom as a function of depth, collecting charge for 5 s at each step. Chamber readings are corrected for polarity effects by taking an average of measurements made at both applied polarities (300 V for cylindrical chambers and 100 V for plane-parallel chambers) and are corrected for ion recombination using previously measured recombination parameters. These corrected chamber readings are normalized to monitor chamber readings throughout this work. Chambers investigated are all NRC secondary standard reference chambers. A set of Farmer-type (graphite-walled, aluminum electrode) cylindrical chambers with nominally identical specifications and very similar (0.6 cm 3 ) collecting volumes is used. Four NE2571 chambers and one NE2505/3 chamber are compared to an additional reference NE2571 chamber that has stability history going back two decades. 22 The NE2505/3 chamber was manufactured earlier (1970s) than the NE2571, and the only difference in the design is that the stem used for the NE2505/3 is not as well guarded. In addition, two PTW Roos and one Scanditronix NACP-02 chamber are compared to an additional Scanditronix NACP-02 chamber.

3 6643 Muir and McEwen: Cylindrical chambers in electron beams B. Measurements in plastic phantom We pointed out in our previous work that using the depthionization method described above to obtain chamber data did not produce useful results in the 4 MeV beam because steep dose gradients result in large uncertainties from positioning. To further investigate relative ion chamber perturbation corrections for cylindrical chambers, measurements are made in a Virtual Waterâ phantom. A hole is bored to accommodate Farmer-type chambers such that the center of the chamber is at the reference depth (0.9 cm) in a 4 MeV beam. The reference depth (d ref = 0.6R cm) is determined from depth-dose measurements in water and scaling the thickness of Virtual Waterâ for water-equivalence with the results of McEwen and Niven. 23 By making minimal changes to the set-up, only substituting one chamber for another, we reduce uncertainties from positioning along the beam axis. The setup for these measurements is shown in Fig. 1. Chambers are preirradiated with 1000 MU, and a set of 200 MU irradiations are carried out while taking integrated charge readings. These measurements are performed at applied voltages of 300 V to correct for polarity effects. 1 Similar to the measurements described in section 2.A., cylindrical Farmer-type chambers with nominally identical specifications are used for these measurements. Three NE2571 chambers, one NE2505/3 chamber, and one IBA FC65-G chamber are compared to the NE2571 chamber that has produced reliable data over two decades. The ratio of chamber readings is then normalized to the same ratio in the 18 MeV beam at the reference depth. We also performed the normalization with results from a cobalt-60 reference field. 3. RESULTS Figure 2 shows the normalized ratio of readings from a set of four NE2571 chambers and an NE2505/3 chamber to those FIG. 2. The variability of relative ion chamber perturbation corrections for graphite-walled chambers with aluminum electrodes and nominally identical specifications as a function of mean electron energy at depth. The y-axis shows the ratio of readings from different NE2571 ion chambers and an NE2505/3 chamber to those from a representative NE2571 reference chamber normalized to the average of the three highest energy points. Open symbols are results from the 18 MeV beam while closed symbols are from 8 MeV. Uncertainties from differences in response over the course of five years and positioning are shown for a representative chamber. [Color figure can be viewed at wileyonlinelibrary.com] from a representative NE2571 reference chamber as a function of mean electron energy at depth. In section 1.B., we pointed out that these normalized ratios can be thought of as relative ion chamber perturbation corrections. We discuss the variability of relative ion chamber perturbation corrections for these chambers in section IV. Figure 3 shows similar results for plane-parallel chambers and demonstrates that the variability of relative perturbation corrections for plane-parallel chambers is no better than that for cylindrical chambers shown in Fig. 2. FIG. 1. The set-up used for Virtual Waterâ measurements. The water phantom behind the Virtual Waterâ phantom, for additional backscatter, can be seen as well as the monitor chambers mounted on the applicator with laboratory hardware and an air temperature sensor. [Color figure can be viewed at wileyonlinelibrary.com] FIG. 3. As in Fig. 2 but for plane-parallel chambers, this time taking the ratio of readings to a reference NACP-02 chamber. Open symbols are results from the 18 MeV beam while closed symbols are from 8 MeV. Uncertainties from differences in response over the course of 5 yr and positioning are shown for a representative chamber. [Color figure can be viewed at wileyonlinelibrary.- com]

4 6644 Muir and McEwen: Cylindrical chambers in electron beams 6644 Table I shows the normalized ratio of chamber readings for cylindrical chambers in the Virtual Waterâ set-up in the 4 MeV beam. This table shows that the variability of relative perturbation factors at the reference depth in the 4 MeV beam is less than about 0.2% for all nominally identical Farmer chambers investigated. Since uncertainty from positioning chambers along the beam axis is minimized the uncertainty associated with these relative measurements is estimated to be 0.2%. 20 These results are normalized with data from the 18 MeV beam as in Figs. 2 and 3, but performing the normalization with data obtained in cobalt-60 does not change the outcome. 4. DISCUSSION In section 1.A., we described how fluence correction factors for cylindrical chambers were determined and subsequently used for protocol calculations of electron beam k Q factors, and pointed out why cylindrical chambers were not recommended for calibration of beams with nominal energy less than 10 MeV. In section III, we provide relative perturbation corrections for a set of cylindrical chambers with nominally identical specifications. We observed that even at E z = 2 MeV (corresponding to the value of E z at the reference depth in a 4 MeV beam) the variation in these relative perturbation corrections is only 0.4%. The results shown above for plane-parallel chambers demonstrate the same level of variability for similar chambers. In addition, measurements made with cylindrical chambers at the reference depth in a 4 MeV beam in a Virtual Waterâ phantom vary by less than about 0.2%. These measurements are for various graphite-walled Farmer-type chambers that employ aluminum electrodes and have similar collecting volume. The different chambers investigated cover a wide time period of manufacturing, from the NE2505/3 #2441 manufactured in the 1970s to the NE2571 #3610 manufactured in Although the perturbation corrections themselves may be up to 5% as pointed out in early fundamental publications, 8 10 these measurements suggest that there is less uncertainty from chamber-to-chamber variations in reference dosimetry data than previously thought. That the chambers were produced over a period of decades, and therefore might be subject to changes in manufacturing, only gives more confidence to our conclusion that there is a lack of variability in perturbation factors for Farmer-type chambers. As a complement to the measurements carried out in this work, we reanalyzed Farmer-chamber data produced for the NPL CIRM-20 report. 24 In that investigation, in-water measurements were limited to energies greater than 10 MeV, but data for each NE2571 was also acquired using the same 90 Sr/ 90 Y check source (type NE2502). We can therefore construct an equivalent to the left side of Eq. (3) where Q is the 90 Sr/ 90 Y check source and Q ecal is the highest energy beam used by NPL in that investigation (18 MeV). The in-air measurement using the check source is not quite the same as an in-water chamber ratio but the very low energy of the beta source (E av < 1 MeV) should amplify chamber-to-chamber variations in response. Figure 2 of McEwen and Taank 25 demonstrates the reproducibility (standard deviation of 0.05%) of chamber measurements in a 90 Sr/ 90 Y check source over the course of a year when repeated set-ups are performed with the same chamber, lending confidence to these measurements. The measurements performed at NPL used a different check source and reproducibility was not quite as good as that demonstrated by McEwen and Taank, 25 but repeated measurements were still obtained at the 0.2% level. We took the ratio of readings from chambers irradiated with the 90 Sr/ 90 Y check source to those from a representative NE2571 chamber, and normalized these measurements with the readings obtained with chambers at the reference depth in water irradiated with an 18 MeV electron beam. These results are presented in Fig. 4 and the quantity of interest is the standard deviation, which is 0.5%, indicating little chamber-to-chamber variation in relative response from high to low electron energy. The small variation demonstrated in Fig. 4 of data acquired under vastly different irradiation geometries supports the findings of the present work. TABLE I. Relative measurements made in a Virtual Waterâ phantom irradiated with a 4 MeV beam for graphite-walled cylindrical chambers with aluminum electrodes and similar collecting volume positioned in a hole bored at the reference depth. The results are normalized to the same ratio in the 18 MeV beam at the reference depth. Chamber Normalized ratio of chamber readings NE2571 # NE2571 # NE2571 # IBA FC65G # NE2505/3 # FIG. 4. The variability of relative ion chamber perturbation corrections for graphite-walled chambers with aluminum electrodes and nominally identical specifications as a function of serial number. The y-axis shows the ratio of readings from different NE2571 ion chambers to those from a representative NE2571 reference chamber measured when irradiated with a 90 Sr/ 90 Y check source normalized to the readings made at the reference depth in water in an 18 MeV beam. Data are reanalyzed from the NPL CIRM-20 report. 24

5 6645 Muir and McEwen: Cylindrical chambers in electron beams 6645 In addition, the results from the NPL electron beam calibration service 24 demonstrate less than 0.15% variation in energy response for 11 Farmer chambers in beams with nominal energy between 10 MeV and 19 MeV. It is difficult to imagine any mechanism that would increase this level of variability dramatically in lower energy beams. On the other hand, Stucki and V or os 26 indicate that the standard deviation in beam quality conversion factors for several PTW Roos and NACP-02 chambers is up to 1.7%. Muir and Rogers 11 show that varying the geometry for Monte Carlo calculations with an NE2571 in an 18 MeV beam results in a deviation less than 0.2%, but small differences in geometry for a PTW Markus chamber can result in up to 1% effects in low-energy beams. To supplement this analysis, we obtained linac output data for electron beams from site visits made by IROC Houston. For site visit calibrations a cylindrical chamber (typically an Exradin A12) is used and compared to the linac calibration at a given site, which could have been based on either a planeparallel or cylindrical chamber. The standard deviation of the linac output determined by IROC Houston from approximately 457 site visits and nearly 4000 beams is around 1.7% and does not vary with beam energy between 5 MeV and 22 MeV. If there was variability from using a cylindrical chamber for calibration then one would expect an increase in standard deviation with decreasing beam energy. That this does not occur strengthens the findings of the present work. This analysis has so far been focused on chamber-tochamber variability but has not yet touched on difficulties associated with chamber positioning along the beam axis resulting in large uncertainties in chamber measurements especially in low-energy beams, where steep dose gradients are present even near the reference depth. 20 However, cylindrical and plane-parallel chambers can be positioned along the beam axis with the same level of accuracy so this factor does not lead to one type being favourable. Reference dosimetry protocols also recommend the use of a cross-calibration procedure where plane-parallel chambers are calibrated against a cylindrical chamber in a high-energy beam and then used for measurements in beams with lower nominal energy. This was because of potentially large (about 4%) variations in wall perturbation factors 12 that would introduce uncertainties in protocol calculations of k Q factors. More recently it has been shown that the variation in measured k Q factors for plane-parallel chambers of the same type is at the 1% level, potentially from improvements in manufacturing. It is more concerning that the long-term stability of plane-parallel chambers is not as good as that for cylindrical chambers and could be affected by environmental conditions ,27 Therefore, to avoid stability issues one would still need to cross-calibrate plane-parallel chambers against stable cylindrical chambers or use some other method of stability monitoring each time they are used. The crosscalibration method complicates the procedure for electron beam dosimetry measurements. It is now possible to directly calculate electron beam k Q factors with confidence with Monte Carlo simulations. 11,28 30 Muir and Rogers 11,29 compared k Q factors for cylindrical and plane-parallel chambers to results from several measurement and Monte Carlo-based publications and observed excellent agreement. We now have high-quality Monte Carlo calculations of electron beam k Q factors for cylindrical chambers and have demonstrated a lack of chamber-to-chamber variability for these chambers in section 3 for relative perturbation corrections at low electron energies. Although the measurements of this work are only conclusive for graphitewalled Farmer-type chambers, there is nothing here that precludes the use of cylindrical chambers for all electron beam reference dosimetry measurements, even in beams with low energy. The analysis here using site visit data from IROC Houston, that would include measurements with cylindrical chambers that use different wall materials, supports the findings of this work. 5. CONCLUSIONS We discussed several issues related to the use of planeparallel chambers for reference dosimetry measurements in electron beams contrary to the recommendations in reference dosimetry protocols. The observations made here are applicable for graphite-walled cylindrical chambers that use aluminum electrodes and have similar collecting volume, so it is difficult to draw conclusions about other cylindrical chamber types. However, since it was potential variability of the fluence correction factor (i.e., variability in cavity dimensions) that led to the recommendation not to use cylindrical chambers, the wall and electrode material may not be relevant. In addition, we demonstrated a lack of variability with beam energy for output results from IROC Houston site visits that would have included several different chamber types, including cylindrical chambers that use different wall materials, and this analysis confirms the findings in this work. Therefore, these results support the suitability of the use of cylindrical chambers even in low-energy electron beams, which would simplify and improve the accuracy of the electron beam calibration procedure. This work will be important for updating protocols for reference dosimetry of linac electron beams. ACKNOWLEDGMENTS The authors thank Graham Bass of the National Physical Laboratory, in the UK, for providing the original NE2571 data from their investigations. The authors also thank David Followill of IROC Houston for providing site visit data. CONFLICT OF INTEREST The authors declare that they do not have any conflict of interest. a) Author to whom correspondence should be addressed. Electronic mail: Bryan.Muir@nrc-cnrc.gc.ca.

6 6646 Muir and McEwen: Cylindrical chambers in electron beams 6646 REFERENCES 1. Almond PR, Biggs PJ, Coursey BM, et al. AAPM's TG 51 protocol for clinical reference dosimetry of high-energy photon and electron beams. Med Phys. 1999;26: IAEA. Absorbed dose determination in external beam radiotherapy: an international code of practice for dosimetry based on standards of absorbed dose to water V12, Technical Report Series Thwaites DI, DuSautoy AR, Jordan T, et al. The IPEMB code of practice for electron dosimetry for radiotherapy beams of initial energy from 4 to 25 MeV based on an absorbed dose to water calibration. Phys Med Biol. 2003;48: Rosenow E, Kasten G, Thienel T. Comparison of three parallel plate ionization chambers for high energy electron dosimetry. In: IAEA, ed. Measurement Assurance in Dosimetry. Vienna: IAEA; 1994: Stewart KJ, Seuntjens JP. Comparing calibration methods of electron beams using plane-parallel chambers with absorbed-dose to water based protocols. Med Phys. 2002;29: Zink K, Czarnecki D, Looe H, Voigts-Rhetz P, Harder D. Monte Carlo study of the depth-dependent fluence perturbation in parallel-plate ionization chambers in electron beams. Med Phys. 2014;41: Buckley LA, Rogers DWO. Wall correction factors, P wall, for parallelplate ionization chambers. Med Phys. 2006;33: Johansson KA, Mattsson LO, Lindborg L, Svensson H. Absorbed-dose determination with ionization chambers in electron and photon beams having energies between 1 and 50 MeV. In: IAEA, ed. IAEA Symposium Proceedings. Vienna: IAEA SM 222/35, IAEA; 1978: Wittk amper FW, Thierens H, Van der Plaetsen A, de Wagter C, Mijnheer BJ. Perturbation correction factors for some ionization chambers commonly applied in electron beams. Phys Med Biol. 1991;36: Van der Plaetsen A, Seuntjens J, Thierens H, Vynckier S. Verification of absorbed doses determined with thimble and parallel-plate ionization chambers in clinical electron beams using ferrous sulphate dosimetry. Med Phys. 1994;21: Muir BR, Rogers DWO. Monte Carlo calculations for reference dosimetry of electron beams with the PTW Roos and NE2571 ion chambers. Med Phys. 2013;40: Kosunen A, J arvinen H, Sipil a P. Optimum calibration of NACP type plane-parallel ionization chambers for absorbed dose determinations in low energy electron beams. In: IAEA, ed. Proceeding of Symposium on Measurement Assurance in Dosimetry. Vienna: IAEA SM 330/419, IAEA; 1994: Christ G, Dohm OS, Bruggmoser G, Sch ule E. The use of plane-parallel chambers in electron dosimetry without any cross-calibration. Phys Med Biol. 2002;47:N121 N Kapsch RP, Gomola I. Beam quality correction factors for plane parallel chambers in photon beams, IAEA E2-CN-182. In: IAEA, ed. Book of Extended Synopses for Symp. on Standards, Applications and Quality Assurance in Medical Radiation Dosimetry. Vienna: IAEA; 2010: Muir BR, McEwen MR, Rogers DWO. Beam quality conversion factors for parallel-plate ionization chambers in MV photon beams. Med Phys. 2012;39: Bass G, Thomas R, Pearce J. The calibration of parallel-plate electron ionization chambers at NPL for use with the IPEM 2003 code of practice: summary data. Phys Med Biol. 2009;54:N115 N Muir B. Ion chamber absorbed dose calibration coefficients, N D, w, measured at ADCLs: Distribution analysis and stability. Med Phys. 2015;42: Muir BR, Cojocaru CD, McEwen MR, Ross CK. Electron beam water calorimetry measurements to obtain beam quality conversion factors. Med Phys. 2017;44: Muir B, Culberson W, Davis S, et al. Insight gained from responses to surveys on reference dosimetry practices. J App Clin Med Phys. 2017;18: Muir BR, McEwen MR, Rogers DWO. Determination of relative ion chamber calibrations coefficients from depth-ionization measurements in clinical electron beams. Phys Med Biol. 2014;59: Thwaites DI, Burns DT, Klevenhagen SC, Nahum AE, Pitchford WG. The IPEMB code of practice for electron dosimetry for radiotherapy beams of initial energy from 2 to 50 MeV based on air kerma calibration. Phys Med Biol.1996;41: Muir BR, McEwen MR, Rogers DWO. Measured and Monte Carlo calculated k Q factors: accuracy and comparison. Med Phys. 2011;38: McEwen MR, Niven D. Characterization of the phantom material virtual water in high-energy photon and electron beams. Med Phys. 2006;33: McEwen MR, Williams AJ, DuSautoy AR. Trial calibrations of therapy level electron beam ionisation chambers in terms of absorbed dose to water. NPL Report CIRM 20, NPL, Teddington, UK; McEwen MR, Taank J. Examining the influence of humidity on reference ionization chamber performance. Med Phys. 2017;44: Stucki G, V or os S. Experimental k Q,Q0 electron beam quality correction factors for the types NACP02 and PTW34001 plane-parallel chambers. Proceed. of Absorbed Dose and Air Kerma Primary Standards Workshop, LNHB, Paris, Palm A, Mattsson O, Andreo P. Calibration of plane-parallel chambers and determination of p wall for the NACP and Roos chambers for 60 Co c- ray beams. Phys Med Biol. 2000;45: Sempau J, Andreo P, Aldana J, Mazurier J, Salvat F. Electron beam quality correction factors for plane-parallel ionization chambers: Monte Carlo calculations using the PENELOPE system. Phys Med Biol. 2004;49: Muir BR, Rogers DWO. Monte Carlo calculations of electron beam quality conversion factors for several ion chamber types. Med. Phys. 2014;41: Zink K, Wulff J. Beam quality corrections for parallel-plate ion chambers in electron reference dosimetry. Phys Med Biol. 2012;57: