Methods for beam data acquisition offered by a mini-phantom

Mini-phantoms are an important tool for measurement of basic head scatter parameters in high-energy photon beams, and recently they have also been used for beam quality specification. Therefore the feasibility and reliability of basic beam parameter acquisition using only a mini-phantom is checked in 6, 18 and 25 MV photon beams. These parameters include head scatter correction factors, phantom scatter correction factors, total scatter correction factors, wedge factors, off-axis ratios, as well as beam attenuation coefficients and beam hardening coefficients. In order to specify beam quality variations and beam quality modifications by a wedge, two different methods are compared: the first method uses a constant source to chamber distance of 1 m, the second method refers to narrow beam geometry. Mu values derived with two different beam quality specification methods show a systematic deviation. However, relative variations of the attenuation coefficient within the beam and the associated beam quality modifications observed with the two methods show good agreement in open and wedged beams. Phantom scatter correction factors are calculated from measured head scatter correction factors and total scatter correction factors as well as from attenuation coefficients. Measured and calculated phantom scatter correction factors agree within 1% with the values given in literature. For 18 and 25 MV photon beam, wedge factors measured in water or in the mini-phantom agree within 0.5%, but maximum deviations of approximately 1.5% are observed at 6 MV for the largest field sizes. It is demonstrated that the determination of several beam data related to full scatter conditions does not necessarily require the availability of a full scatter phantom. The mini-phantom is a reliable but very cheap and simple tool. It offers versatile possibilities to measure, check and verify basic beam parameters in high-energy photon beams.     

Optimization of photon beam flatness for radiation therapy

In this work, we investigate the relation between lateral fluence/dose distributions and photon beam uniformity, possibly identifying ways to improve these characteristics. The calculations included treatment head scatter properties associated with three common types of linear accelerators in order to study their impact on the results. For 6 and 18 MV photon beams the lateral fluence distributions were optimized with respect to the resulting calculated flatness, as defined by the International Electrotechnical Commission (IEC), at 10 cm depth in six different field sizes. The limits proposed by IEC for maximum dose ratios ('horns') at the depth of dose maximum have also been accounted for in the optimization procedure. The conclusion was that typical head scatter variations among different types of linear accelerators have a very limited effect on the optimized results, which implies that the existing differences in measured off-axis dose distributions are related to non-equivalent optimization objectives. Finally, a comparison between the theoretically optimized lateral dose distributions and corresponding dose measurements for the three investigated accelerator types was performed. Although the measured data generally fall within the IEC requirements the optimized distributions show better results overall for the evaluated uniformity parameters, indicating that there is room for improved flatness performance in clinical photon beams.     

Dosimetric properties of photon beams from a flattening filter free clinical accelerator

Basic dosimetric properties of 6 MV and 18 MV photon beams from a Varian Clinac 21EX accelerator operating without the flattening filter have been measured. These include dose rate data, depth dose dependencies and lateral profiles in a water phantom, total scatter factors and transmission factors of a multileaf collimator. The data are reviewed and compared with measurements for the flattened beams. The unflattened beams have the following: a higher dose rate by factors of 2.3 (6 MV) and 5.5 (18 MV) on the central axis; lower out-of-field dose due to reduced head scatter and softer spectra; less variation of the total scatter factor with field size; and less variation of the shape of lateral dose profiles with depth. The findings suggest that with a flattening filter free accelerator better radiation treatments can be developed, with shorter delivery times and lower doses to normal tissues and organs.     

On empirical methods to determine scatter factors for irregular MLC shaped beams

Multileaf collimators (MLCs) are in clinical use for more than a decade and are a well accepted tool in radiotherapy. For almost each MLC design different empirical or semianalytical methods have been presented for calculating output ratios in air for irregularly shaped beams. However, until now no clear recommendations have been given on how to handle irregular fields shaped by multileaf collimators for independent monitor unit (MU) verification. The present article compares different empirical methods, which have been proposed for independent MU verification, to determine (1) output ratios in air (Sc) and (2) phantom scatter factors (Sp) for irregular MLC shaped fields. Ten dedicated field shapes were applied to five different types of MLCs (Elekta, Siemens, Varian, Scanditronix, General Electric). All calculations based on empirical relations were compared with measurements and with calculations performed by a treatment planning system with a fluence based algorithm. For most irregular MLC shaped beams output ratios in air could be adequately modeled with an accuracy of about 1%-1.5% applying a method based on the open field aperture defined by the leaf and jaw setting combined with the equivalent square formula suggested by Vadash and Bj?rngard [P. Vadash and B. E. Bj?rngard, Med. Phys. 20, 733-734 (1993)]. The accuracy of this approach strongly depends on the inherent head scatter characteristics of the accelerator in use and on the irregular field under consideration. Deviations of up to 3% were obtained for fields where leaves obscure central parts of the flattening filter. Simple equivalent square methods for Sp calculations in irregular fields did not provide acceptable results (deviations mostly >3%). Sp values derived from Clarkson integration, based on published tables of phantom scatter correction factors, showed the same accuracy level as calculations performed using a pencil beam algorithm of a treatment planning system (in a homogeneous media). The separation of head scatter and phantom scatter contributions is strongly recommended for irregular MLC shaped beams as both contributions have different factors of influence. With rather simple methods Sc and Sp can be determined for independent MU calculation with an accuracy better than 1.5% for most clinical situations encountered in conformal radiotherapy.     

A 4 MV flattening filter-free beam: commissioning and application to conformal therapy and volumetric modulated arc therapy

Recent studies have indicated that radiotherapy treatments undertaken on a flattening filter-free (FFF) linear accelerator have a number of advantages over treatments undertaken on a conventional linear accelerator. In addition, 4 MV photon beams may give improved isodose coverage for some treatment volumes at air/tissue interfaces, compared to when utilizing the clinical standard of 6 MV photons. In order to investigate these benefits, FFF beams were established on an Elekta Beam Modulator linear accelerator for 4 MV photons. Commissioning beam data were obtained for open and wedged fields. The measured data were then imported into a treatment planning system and a beam model was commissioned. The beam model was optimized to improve dose calculations at shallow, clinically relevant depths. Following verification, the beam model was utilized in a treatment planning study, including volumetric modulated arc therapy, for a selection of lung, breast/chest wall and larynx patients. Increased dose rates of around 800 MU min(-1) were recorded for open fields (relative to 320 MU min(-1) for filtered open fields) and reduced head scatter was inferred from output factor measurements. Good agreement between planned and delivered dose was observed in verification of treatment plans. The planning study indicated that with a FFF beam, equivalent (and in some cases improved) isodose profiles could be achieved for small lung and larynx treatment volumes relative to 4 MV filtered treatments. Furthermore, FFF treatments with wedges could be replicated using open fields together with an 'effective wedge' technique and isocentre shift. Clinical feasibility of a FFF beam was therefore demonstrated, with beam modelling, treatment planning and verification being successfully accomplished.     

Internal scatter, the unavoidable major component of the peripheral dose in photon-beam radiotherapy

In clinical photon beams, the dose outside the geometrical field limits is produced by photons originating from (i) head leakage, (ii) scattering at the beam collimators and the flattening filter (head scatter) and (iii) scattering from the directly irradiated region of the patient or phantom (internal scatter). While the first two components can be modified, e.g. by reinforcement of shielding components or by re-modeling the filter system, internal scatter remains an unavoidable contributor to the peripheral dose. Its relative magnitude compared to the other components, its numerical variation with beam energy, field size and off-axis distance as well as its spectral distribution are evaluated in this study. We applied a detailed Monte Carlo (MC) model of our 6/15 MV Siemens Primus linear accelerator beam head, provided with ideal head leakage shielding conditions (multi-leaf collimator without gaps) to assess the head scatter contribution. Experimental values obtained under real shielding conditions were used to evaluate the head leakage contribution. It was found that the MC-computed internal scatter doses agree with the results of our previous measurements, that internal scatter is the major contributor to the peripheral dose in the near periphery while head leakage prevails in the far periphery, and that the lateral decline of the internal scatter dose can be represented by the sum of two exponentials, with an asymptotic tenth value of 18 to 19 cm. Internal scatter peripheral doses from various elementary beams are additive, so that their sum increases approximately in proportion with field size. The ratio between normalized internal scatter doses at 6 and 15 MV is approximately 2:1. The energy fluence spectra of the internal scatter component at all points of interest outside the field have peaks near 500 keV. The fact that the energy-shifted internal scatter constitutes the major contributor to the dose in the near periphery has a general bearing for dosimetry, i.e. for energy-dependent detector responses and dose conversion factors, for the relative biological effectiveness and for second primary malignancy risk estimates in the peripheral region.     

Modeling the extrafocal radiation and monitor chamber backscatter for photon beam dose calculation

A simple analytical approach has been developed to model extrafocal radiation and monitor chamber backscatter for clinical photon beams. Model parameters for both the extrafocal source and monitor chamber backscatter are determined simultaneously using conventional measured data, i.e., in-air output factors for square and rectangular fields defined by the photon jaws. The model has been applied to 6 MV and 15 MV photon beams produced by a Varian Clinac 2300C/D accelerator. Contributions to the in-air output factor from the extrafocal radiation and monitor chamber backscatter, as predicted by the model, are in good agreement with the measurements. The model can be used to calculate the in-air output factors analytically, with an accuracy of 0.2% for symmetric or asymmetric rectangular fields defined by jaws when the calculation point is at the isocenter and 0.5% when the calculation point is at an extended SSD. For MLC-defined fields, with the jaws at the recommended positions, calculated in-air output factors agree with the measured data to within 0.3% at the isocenter and 0.7% at off-axis positions. The model has been incorporated into a Monte Carlo dose algorithm to calculate the absolute dose distributions in patients or phantoms. For three MLC-defined irregular fields (triangle shape, C-shape, and L-shape), the calculations agree with the measurements to about 1% even for points at off-axis positions. The model will be particularly useful for IMRT dose calculations because it accurately predicts beam output and penumbra dose.     

Photonuclear dose calculations for high-energy photon beams from Siemens and Varian linacs

The dose from photon-induced nuclear particles (neutrons, protons, and alpha particles) generated by high-energy photon beams from medical linacs is investigated. Monte Carlo calculations using the MCNPX code are performed for three different photon beams from two different machines: Siemens 18 MV, Varian 15 MV, and Varian 18 MV. The linac head components are simulated in detail. The dose distributions from photons, neutrons, protons, and alpha particles are calculated in a tissue-equivalent phantom. Neutrons are generated in both the linac head and the phantom. This study includes (a) field size effects, (b) off-axis dose profiles, (c) neutron contribution from the linac head, (d) dose contribution from capture gamma rays, (e) phantom heterogeneity effects, and (f) effects of primary electron energy shift. Results are presented in terms of absolute dose distributions and also in terms of DER (dose equivalent ratio). The DER is the maximum dose from the particle (neutron, proton, or alpha) divided by the maximum photon dose, multiplied by the particle quality factor and the modulation scaling factor. The total DER including neutrons, protons, and alphas is about 0.66 cSv/Gy for the Siemens 18 MV beam (10 cm x 10 cm). The neutron DER decreases with decreasing field size while the proton (or alpha) DER does not vary significantly except for the 1 cm x 1 cm field. Both Varian beams (15 and 18 MV) produce more neutrons, protons, and alphas particles than the Siemens 18 MV beam. This is mainly due to their higher primary electron energies: 15 and 18.3 MeV, respectively, vs 14 MeV for the Siemens 18 MV beam. For all beams, neutrons contribute more than 75% of the total DER, except for the 1 cm x 1 cm field (approximately 50%). The total DER is 1.52 and 2.86 cSv/Gy for the 15 and 18 MV Varian beams (10 cm x 10 cm), respectively. Media with relatively high-Z elements like bone may increase the dose from heavy charged particles by a factor 4. The total DER is sensitive to primary electron energy shift. A Siemens 18 MV beam with 15 MeV (instead of 14 MeV) primary electrons would increase by 40% the neutron DER and by 210% the proton + alpha DER. Comparisons with measurements (neutron yields from different materials and neutron dose equivalent) are also presented. Using the NCRP risk assessment method, we found that the dose equivalent from leakage neutrons (at 50-cm off-axis distance) represent 1.1, 1.1, and 2.0% likelihood of fatal secondary cancer for a 70 Gy treatment delivered by the Siemens 18 MV, Varian 15 MV, and Varian 18 MV beams, respectively.     

Derivation of the distribution of extrafocal radiation for head scatter factor calculation

Head scatter factors for high energy photon beams from linear accelerators can be modeled using a two-source model consisting of focal and extrafocal radiation. The focal radiation can be approximated as a point source, and the distribution of the extrafocal radiation is a two-dimensional (2D) radial symmetric function. Various methods, including analytical, Monte Carlo, and empirical trial functions, have been used to determine the radial symmetric function of extrafocal radiation distribution. This article describes a method for directly determining the extrafocal radiation distribution without assuming any empirical trial function. The extrafocal radiation distribution is determined with measured head scatter factors for rectangular fields defined by the lower jaw (X) fixed at 40 cm and the upper jaw (Y) varying from 3 to 40 cm. The derivatives of the measured head scatter factors, with respect to the Y jaw position projected in the plane of extrafocal radiation, are proportional to the one-dimensional (1D) projection (also called the line spread function) of the extrafocal radiation distribution. Two methods are used to solve the radial function of extrafocal radiation from the 1D projection. The first method uses a 2D filtered backprojection algorithm, originally developed for parallel beam computed tomography reconstruction, to directly derive the radial dependence of the extrafocal radiation distribution. The method has been applied to 6 and 18 MV photon beams from a Siemens linear accelerator and has been tested by comparing measured and calculated head scatter factors for square and rectangular fields. The second method uses a Fourier transform followed by a Fourier-Bessel transform to solve the problem. The distributions of extrafocal radiation derived from these two methods are virtually identical.     

Spectral reconstruction of high energy photon beams for kernel based dose calculations

A kernel-based dose computation method with finite-size pencil beams (FSPBs) requires knowledge of the photon spectrum. Published methods of indirect spectral measurements using transmission measurements through beam attenuators use mathematical fits with a large number of parameters and constraints. In this study, we examine a simple strategy for fitting transmission data that models important physical characteristics of photon beams produced in clinical linear accelerators. The shape of an unattenuated bremsstrahlung spectrum is known, varying linearly from a maximum at zero energy to a value of zero at a maximum energy. This unattenuated spectrum is altered primarily by absorption of low energy photons by the flattening filter, causing the true spectrum to roll off to zero at low photon energies. A fitting equation models this behavior and has these advantages over previous methods: (1) the equation describes the shape of a bremsstrahlung spectrum based on physical expectations; and (2) only three fit parameters are required with a single constraint. Results for 4 MV and 6 MV accelerators for central axis and off-axis beams show good agreement with the maximum, average and modal energies for known spectra. Previously published models, representations of beam fluence (energy fluence, dN/dE), experimental methods, and the fitting process are discussed.     

Wedge factor dependence with depth, field size, and nominal distance--a general computational rule

We have investigated the dependence of the wedge factors with field size, depth, nominal, and extended distances for 4, 6, 18, and 24 MV photon beams. Analysis of the experimental data suggests a general linear dependence of the wedge factors with field size and depth. The study shows that changes in wedge factors are insignificant (< or = +/-1.0%) with respect to measurements at nominal SSD, SAD, or extended SSD. This independence of the wedge factors on source-to-surface distance was studied for different photon energies (4-24 MV) and for different attenuating wedges (external and internal wedges). For clinical applications, an algorithm is presented to calculate the wedge factor dependence with field size and depth. The new algorithm has been successfully implemented to replace wedge look-up tables for dose and MU calculations in PRISM 1.2 treatment planning system used in our department.     

Properties of unflattened photon beams shaped by a multileaf collimator

Several studies have shown that removal of the flattening filter from the treatment head of a clinical accelerator increases the dose rate and changes the lateral profile in radiation therapy with photons. However, the multileaf collimator (MLC) used to shape the field was not taken into consideration in these studies. We therefore investigated the effect of the MLC on flattened and unflattened beams. To do this, we performed measurements on a Varian Clinac 21EX and MCNPX Monte Carlo simulations to analyze the physical properties of the photon beam. We compared lateral profiles, depth dose curves, MLC leakages, and total scatter factors for two energies (6 and 18 MV) of MLC-shaped fields and jaw-shaped fields. Our study showed that flattening filter-free beams shaped by a MLC differ from the jaw-shaped beams. Similar differences were also observed for flattened beams. Although both collimating methods produced identical depth dose curves, the penumbra size and the MLC leakage were reduced in the softer, unflattened beam and the total scatter factors showed a smaller field size dependence.     

Photon scatter kernels for intensity modulating radiation therapy filters.

The most important beam property while optimizing photon therapy is the ability to modulate the intensity of the beam. The use of photon absorbers for intensity modulation of beam profiles requires special attention to be paid to the alteration of beam properties due to scatter and spectral changes, in addition to the desired intensity modulation. In this study the influence of photon scatter in high-density filters irradiated with very narrow photon pencil beams was investigated. A simple analytical relation is developed to quantify the contribution by scattered photons. A scatter kernel was derived by convolving the first Compton scatter distribution with an approximate expression for the second-order scattered photons. The calculations were validated experimentally with film dosimetry and also by using Monte Carlo simulations. Results show that the difference in photon scatter estimation by different methods is relatively small when higher order scattering is accounted for. At 6 MV x-rays the agreement is slightly better than that for 18 MV x-rays results. The simple relation presented in this paper can be used to account for the scattered photon contribution in filter optimization codes to deliver biologically or physically optimized intensity modulated treatments.     

Analytical representation of enhanced dynamic wedge factors for symmetric and asymmetric photon fields

The Enhanced Dynamic Wedge (EDW) presents many advantages over the physical wedge. However, in order to calculate monitor units (MUs) necessary to deliver a certain dose at a certain point, EDW factors (EDWFs) need to be determined. In this work, based on analysis of the golden segmented treatment table (GSTT) and the MU fraction model, an empirical analytic formula has been developed to calculate EDW factors for symmetric and asymmetric fields. This formalism is an extension of the MU fraction model. However in comparison with previous studies [J. P. Gibbons, Med. Phys. 25, 1411-1418 (1998) and M. Miften et al., Med. Dosim. 25, 81-86 (2000)], this formula is simpler, and easier to use. It is applicable to EDW fields of different sizes, wedge angles and different photon energies. For 6 and 18 MV beams from a Varian 21EX accelerator with 7 EDW angles (Varian Oncology Systems, Palo Alto, CA), more than 250 measured EDWFs for symmetric and asymmetric fields with different off-axis distances and field sizes were compared with model calculations. Results show that 80% and 98% of calculated EDWFs match corresponding measured values to within 0.5% and 1.0%, respectively, the maximum deviation being 1.3%.     

Monitor unit calculations for wedged asymmetric photon beams

Algorithms for calculating monitor units (MUs) in wedged asymmetric high-energy photon beams as implemented in treatment planning systems have their limitations. Therefore an independent method for MU calculation is necessary. The aim of this study was to develop an empirical method to determine MUs for points at the centre of wedged fields, asymmetric in two directions. The method is based on the determination of an off-axis factor (OAF) that corrects for the difference in dose between wedged asymmetric and wedged symmetric beams with the same field size. Measurements were performed in a water phantom irradiated with 6 and 18 MV photon beams produced by Elekta accelerators, which are fitted with an internal motorized wedge that has a complex shape. The OAF perpendicular to the wedge direction changed significantly with depth for the 18 MV beam. Dose values measured for a set of 18 test cases were compared with those calculated with our method. The maximum difference found was 6.5% and in 15 cases this figure was smaller than 2.0%. The analytical method of Khan and the empirical method of Georg were also tested and showed errors up to 12.8%. It can be concluded that our simple formalism is able to calculate MUs in wedged asymmetric fields with an acceptable accuracy in most clinical situations.     

An empirical model for megavoltage x-ray output factors

An empirical model of the factors that determine the central axis dose at 10 cm depth in water for 4 MV, 6 MV and 18 MV photon beams is presented. Backscattering from the variable collimators into the dose monitoring ionization chamber can cause a variation of -0.5% to +1.8% in the dose per monitor unit in accelerators with an electron facility. Forward emission towards the isocentre from the beam flattening filter and upper collimators is more dependent on the position of the upper variable collimator blades than the lower blades, so that they are not interchangeable in determining output factors, which can differ by up to 2%. The model includes the product of the monitor backscatter factor, normalized phantom scatter factor, normalized head scatter factor and inverse square law, corrected for the displacement of the virtual x-ray focus from the target. It can predict the dose to -/+0.83% for 4 MV, -/+0.80% for 6 MV photons and -/+0.82% for 18 MV photons. The normalized head scatter factor is a second-order polynomial of the modified equivalent square collimator, whose coefficients do not vary significantly with x-ray energy. The model was tested by comparison with independent measurements of output factor and generally agreed to around 1%.     

Two-effective-source method for the calculation of in-air output at various source-to-detector distances in wedged fields.

A simple algorithm was developed for calculation of the in-air output at various source-to-detector distances (SDDs) on the central axis for wedged fields. In the algorithm we dealt independently with two effective sources, one for head scatter and the other for wedge scatter. Varian 2100C with 18 and 8 MV photon beams was used to examine this algorithm. The effective source position for head scatter for wedged fields was assumed to be the same as that for open fields, and the effective source position for wedge scatter was assumed to be a certain distance upstream from the physical location of the wedge. The shift of the effective source for wedge scatter, w, was found to be independent of field size. Moreover, we observed no systematic dependency of w on wedge angle or beam energy. One value, w = 5.5 cm, provided less than 1% difference in in-air outputs through the whole experimental range, i.e., 6 x 6 to 20 x 20 cm2 field size (15 x 20 cm2 for 60 degrees wedge), 15 degrees-60 degrees wedge angle, 80-130 cm SDD, and both 18 and 8 MV photon beams. This algorithm can handle the case in which use of a tertiary collimator with an external wedge makes the field size for the determination of wedge scatter different from that for head scatter. In this case, without the two-effective-source method, the maximum of 4.7% and 2.6% difference can be given by the inverse square method and one-effective-source method in a 45 degrees wedged field with 18 MV. Differences can be larger for thicker wedges. Enhanced dynamic wedge (EDW) fields were also examined. It was found that no second effective source is required for EDW fields.     

A Monte Carlo study of a flattening filter-free linear accelerator verified with measurements

A Monte Carlo model of an Elekta Precise linear accelerator has been built and verified by measured data for a 6 and 10 MV photon beam running with and without a flattening filter in the beam line. In this study the flattening filter was replaced with a 6 mm thick copper plate, provided by the linac vendor, in order to stabilize the beam. Several studies have shown that removal of the filter improves some properties of the photon beam, which could be beneficial for radiotherapy treatments. The investigated characteristics of this new beam included output, spectra, mean energy, half value layer and the origin of scattered photons. The results showed an increased dose output per initial electron at the central axis of 1.76 and 2.66 for the 6 and 10 MV beams, respectively. The number of scattered photons from the accelerator head was reduced by (31.7 ± 0.03)% (1 SD) for the 6 MV beam and (47.6 ± 0.02)% for the 10 MV beam. The photon energy spectrum of the unflattened beam was softer compared to a conventional beam and did not vary significantly with the off-axis distance, even for the largest field size (0-20 cm off-axis).     

Ionization chamber dosimetry of small photon fields: a Monte Carlo study on stopping-power ratios for radiosurgery and IMRT beams

Absolute dosimetry with ionization chambers of the narrow photon fields used in stereotactic techniques and IMRT beamlets is constrained by lack of electron equilibrium in the radiation field. It is questionable that stopping-power ratio in dosimetry protocols, obtained for broad photon beams and quasi-electron equilibrium conditions, can be used in the dosimetry of narrow fields while keeping the uncertainty at the same level as for the broad beams used in accelerator calibrations. Monte Carlo simulations have been performed for two 6 MV clinical accelerators (Elekta SL-18 and Siemens Mevatron Primus), equipped with radiosurgery applicators and MLC. Narrow circular and Z-shaped on-axis and off-axis fields, as well as broad IMRT configured beams, have been simulated together with reference 10 x 10 cm2 beams. Phase-space data have been used to generate 3D dose distributions which have been compared satisfactorily with experimental profiles (ion chamber, diodes and film). Photon and electron spectra at various depths in water have been calculated, followed by Spencer-Attix (delta = 10 keV) stopping-power ratio calculations which have been compared to those used in the IAEA TRS-398 code of practice. For water/air and PMMA/air stopping-power ratios, agreements within 0.1% have been obtained for the 10 x 10 cm2 fields. For radiosurgery applicators and narrow MLC beams, the calculated s(w,air) values agree with the reference within +/-0.3%, well within the estimated standard uncertainty of the reference stopping-power ratios (0.5%). Ionization chamber dosimetry of narrow beams at the photon qualities used in this work (6 MV) can therefore be based on stopping-power ratios data in dosimetry protocols. For a modulated 6 MV broad beam used in clinical IMRT, s(w,air) agrees within 0.1% with the value for 10 x 10 cm2, confirming that at low energies IMRT absolute dosimetry can also be based on data for open reference fields. At higher energies (24 MV) the difference in s(w,air) was up to 1.1%, indicating that the use of protocol data for narrow beams in such cases is less accurate than at low energies, and detailed calculations of the dosimetry parameters involved should be performed if similar accuracy to that of 6 MV is sought.