[en] Clinical proton beam therapy has known an explosive growth in the last decennium, a growth that has not ended yet. A huge uniformity in clinical (routine) proton beam dosimetry has been established by two dosimetry protocols (ECHED and AAPM). In spite of this, the influence of chamber dependent parameter on the ionisation chamber dosimetry, has not been cleared out profoundly. A comparative study between water calorimetry using the calorimeter of Gent and ionometry applying the ECHED protocol showed a discrepancy of 2.6% in dose to water response for 85 MeV protons. To evaluate if, and to what extent this difference can be attributed to ion chamber dependent parameters, Monte Carlo (MC) calculations are indispensable to support and to understand measurements. Indications for the existence of chamber dependent corrections are found in the above-mentioned study in which the dose response of different ion chamber types showed systematic differences of up to 1.1%. Existing proton MC codes do not allow to simulate complex geometries with different media. Therefore an existing MC code (PTRAN) has been modified. With the original code it is possible to calculate depth dose profiles, spectral energy and radial distributions for pencil beams in water. The code applies Molliere's multiple scattering theory and Vavilov's energy straggling theory. Changes to the code are made in order to be able to make simulations for other materials than water and for complex geometries with more than one medium. The program's accuracy has been tested by making a comparison with measurements and calculations reported in the literature. Preliminary results of the influence of chamber design and chamber materials on dose to water determination will be presented

[en] The importance of water calorimetry in clinical proton beam dosimetry has increased for various reasons: there has been an increasing interest in proton therapy from the radiotherapy world; existing dosimetry protocols for clinical proton beams recommend calorimetry as the primary dosimetry method (i.e., AAPM and ECHED); water has recently become the reference material for dose specification. For the water dose evaluation, the water calorimeter developed in Ghent was used. Ionometry was performed following the ECHED protocol. The study resulted in a calorimetric to ionometric dose ratio of 0.974 ± 0.009. The discrepancy should in our opinion be attributed to the (W/e)_p,air value of 35.2 J/C implemented in the protocol, possibly next to ion chamber dependent effects for which indications are found in the ionometry measurements. (author)

[en] The increased use of small fields in intensity modulated and stereotactic treatments has created the demand for more standardization of dosimetry procedures for these non-reference fields. In addition, treatment units such as GammaKnife, CyberKnife and TomoTherapy cannot establish broad beam reference conditions prescribed in conventional dosimetry. For dynamic, modulated deliveries, the step between dosimetry in the conventional static broad beam reference field and the actual treatment delivery is large and it has been suggested that performing reference dosimetry for an intermediate field may substantially reduce the uncertainty. This paper reviews the problems associated with introducing standard dosimetry procedures for these non-standard fields, proposed solution and a status of data and information needed for providing recommendations for reference dosimetry. (author)

[en] Full text: While reference dosimetry for broad external beams has been well established the increased use of small fields in intensity modulated radiotherapy (IMRT) and stereotactic radiosurgery (SRS) has revealed that there are many remaining questions about the accuracy of small field dosimetry and what constitutes best practice. Various incidents reported recently demonstrate the need for clear guidelines and recommendations for small field dosimetry. Some treatment units, such as GammaKnife, CyberKnife and TomoTherapy, cannot establish the broad beam reference conditions prescribed in conventional dosimetry. Ionisation chambers, most commonly used and recommended in reference dosimetry, are too voluminous for use in small fields causing perturbation corrections to become unmanageably large. For many other detectors, the energy dependence and its influence in going from large to small field dosimetry is not well known. While there are some indications that beam quality is not greatly affected by field size (based on the small variation of the water to air stopping power ratios), it is not a priori obvious how to measure beam quality for small fields especially if a broad beam cannot be established at the same unit. Even the definition of field size itself becomes ambiguous due to the apparent beam widening compared to the collimator setting as a result of overlapping penumbrae or lateral charged particle disequilibrium in small fields. Concerning composite IMRT plans, it is common practice to verify a clinical plan by calculating the dose distributions it will deliver to a homogeneous phantom and verify it by 2D- or 3D-dosimetry methods. However, with the use of more and more complex IMRT plans in order to achieve homogeneous target dose coverage whilst optimally sparing critical organs, dose distributions in homogeneous phantom conditions become increasingly inhomogeneous and accurate dosimetry becomes very difficult if not impossible. This is exemplary of the huge leap between the static fields used for reference and relative dosimetry on the one hand and the way a clinical IMRT treatment is actually delivered on the other hand. There is also an increased tendency of treatment planning systems to be based on fluence calibrations rather than on dosimetric data for a range of field sizes. For those situations reference dosimetry in a composite reference plan would be more relevant. Several national and international working groups have been established in recent years to provide literature review, guidelines and recommendations for small and composite field dosimetry (IPEM, DIN, NCS, IAEA/AAPM...). An international working group on small and composite field dosimetry formed by the IAEA in collaboration with the AAPM, published a proposed formalism extending the recommendations from IAEA TRS-398 to fields that cannot establish conventional reference conditions as well as to composite fields. This formalism introduces the concepts of machine specific or intermediate reference fields for static small fields and plan-class specific reference fields for composite fields, which both deviate from conventional reference fields and bridge the gap with smaller fields and clinical composite fields, respectively. The dosimetry for both types of fields requires dosimeter perturbation correction factors that can be evaluated using Monte Carlo simulations or experiments. This presentation will review the problems associated with small and composite field dosimetry and recent solutions that have been proposed for various of the above mentioned problems. Concerning beam quality measurement for example, several authors have proposed an equivalent field size method combined with generic tabulated data of depth dose characteristics. Others have demonstrated that dose-area-product measurements combined with lateral distributions can offer an alternative. Dosimeter volume averaging can within certain limits be corrected for by high-resolution 2D- or 3D-dosimetry methods such as film and gel dosimetry. Others have proposed abandoning the concept of measuring dose at a point but rather measure dose over a volume coinciding with the detector's sensitive volume, an approach particularly useful for plan-class specific reference fields. The energy dependence of dosimeters like diodes can be dealt with by cross-calibrating them against ion chambers at an intermediate field, small enough to contain negligible head scatter contribution but large enough to exhibit lateral charged particle equilibrium and minimal volume averaging for the ion chamber. The second part of the presentation will deal with the proposed formalism and review the status of data collection from the literature and recent experimental and Monte Carlo work to derive the necessary correction factors for its application. Based on a discussion of uncertainty requirements of small and composite fields the significance of perturbations emerging from those data is evaluated. The general observation is that for most machine specific or intermediate reference fields and plan-class specific reference fields the correction factors are within 1-2% from unity while for relative field output factors in small fields down to 5 mm field size perturbations are still within a manageable limit of 5% for detectors that exhibit sufficiently low volume averaging and are not too much affected by e.g. high-Z metallic electrodes and cable currents. In general these observations indicate that it is well feasible to apply the formalism but that more data are needed for pinning down correction factors within uncertainty levels of a few tenths of a percent

[en] In the last decade, several clinical proton beam therapy facilities have been developed. To satisfy the demand for uniformity in clinical (routine) proton beam dosimetry two dosimetry protocols (ECHED and AAPM) have been published. Both protocols neglect the influence of ion chamber dependent parameters on dose determination in proton beams because of the scatter properties of these beams, although the problem has not been studied thoroughly yet. A comparison between water calorimetry and ionisation chamber dosimetry showed a discrepancy of 2.6% between the former method and ionometry following the ECHED protocol. Possibly, a small part of this difference can be attributed to chamber dependent correction factors. Indications for this possibility are found in ionometry measurements. To allow the simulation of complex geometries with different media necessary for the study of those corrections, an existing proton Monte Carlo code (PTRAN, Berger) has been modified. The original code, that applies Mollire's multiple scattering theory and Vavilov's energy straggling theory, calculates depth dose profiles, energy distributions and radial distributions for pencil beams in water. Comparisons with measurements and calculations reported in the literature are done to test the program's accuracy. Preliminary results of the influence of chamber design and chamber materials on dose to water determination are presented

[en] Monte Carlo calculations of wall and central electrode perturbation correction factors for Farmer type chambers with graphite and A150 walls and graphite and aluminium central electrodes in proton beams due to secondary electrons are performed using EGSnrc Monte Carlo simulations. The wall correction factors exponentially saturate at high energies at about 1.004 for an A150 wall and about 0.985 for a graphite wall. The central electrode correction is unity for a graphite central electrode and saturates exponentially at 0.998 for a 1 mm diameter aluminium central electrode. Experimental data from the literature are in agreement within the standard uncertainties. The calculated data result in an overall perturbation correction factor for a Farmer type chamber with a graphite wall and a central electrode of 0.9965 in high energy clinical proton beams. (author)

[en] Fluence perturbation of secondary electrons from clinical proton beams (50-250 MeV) by thin high- Z planar interfaces was studied with Monte Carlo simulations. Starting from monoenergetic proton pencil beams, proton depth doses and proton fluence spectra were calculated, both in homogeneous water and near thin high-Z interfaces by using the proton transport Monte Carlo code PTRAN. This code was modified extensively to enable modelling of proton transport in non-homogeneous geometries. From the proton fluence spectra in water and in the interface materials, electron generation spectra were calculated analytically and were then used as input for an electron transport calculation with the Monte Carlo code EGS4/PRESTAII to obtain electron doses and electron fluence spectra. The interface materials used in the study were graphite, Al, Ti, Cu, Sn and Au. We found significant electron fluence perturbations on both sides of the planar interfaces, resulting in an electron dose increase upstream and a decrease downstream from the interfaces, with the magnitude of the effect depending strongly on the atomic number of the interface. For the most extreme case studied, 250 MeV protons and a gold interface, we obtained an electron dose increase of 41% upstream of the interface and a decrease of 15% downstream with both perturbations having a spatial extent of about 700 ^μm. The total dose perturbation due to this effect amounts to a 5% increase upstream and a 2% decrease downstream. A detailed analysis of dose and fluence perturbation is presented for a wide range of materials and proton energies. (author)

[en] Dosimetry of proton therapy has reached a reasonable level of consistency with the introduction of absorbed dose based protocols by the IAEA and the ICRU. Both recommendations adopted the assumption from earlier recommendations that ion chamber perturbation factors in proton beams are unity for all ionisation chamber types. However, early experiments indicated a difference in response of up to 2% between ion chambers with a graphite wall and those with an A150 tissue equivalent wall. A later experiment demonstrated that Farmer type chambers with graphite and A150 walls exhibit on average a difference in response of 0.5%. Monte Carlo simulations showed that secondary electrons are responsible for this effect but the uncertainty of those simulations was rather large due to limitations in available computation time and they were performed with EGS4. Another potential issue is the central electrode effect of chambers with an aluminium central electrode. Two experimental papers indicate that the perturbation factor due to the aluminium electrode is unity within the experimental uncertainties. No Monte Carlo simulations have been reported on this. It is worth to note that in high-energy electron beams, the presently accepted value of the aluminium central electrode perturbation factor is 0.998. Method: Monte Carlo simulations: Simulations of electron transport were performed assuming that the proton spectrum over the geometry of the cavity geometry does not change. This can be retrospectively justified if the resulting perturbation factors are only slowly varying as a function of energy. The modelled geometry is a cylinder with length 24 mm, diameter 6.4 mm and wall thickness of 0.38 mm, i.e. dimensions close to that of Farmer type chambers. Both graphite and A150 as wall material is modelled. The central electrode was modelled as a 1 mm diameter and 20 mm long cylinder at the centre of the cavity protruding from its base. Electron recoil spectra in the cavity, central electrode, wall and surrounding water are calculated from the Bhabba cross section and based on the electron density in each medium. Electron transport through the geometry is simulated using DOSRZnrc for either isotropic internal electron sources or more realistic sources accounting for a simplified angle-energy double-differential distribution of recoil electrons (for which a modification to the code was required). These can be considered as two extremes. Results: The resulting wall and central electrode perturbation factors as a function of incident proton energy are shown. The wall perturbations are energy dependent and increase from unity to about 1.004 for an A150-walled Farmer chamber and decrease from unity to about 0.998 for a graphite walled Farmer chamber. This is in good agreement with the experimental ratio of about 1.005 observed for the same geometries in an 75 MeV proton beam. The central electrode perturbation factor for a 1 mm diameter aluminium electrode in a Farmer decreases from unity to about 0.998. This is consistent with experimental data although the experimental uncertainties do not enable to resolve the 0.2% effect

[en] In the past two decades, the water calorimetry technique for determination of absorbed dose to water in several types of radiation beams has moved significantly closer to being a recognized method. In this paper we summarize the constructional details of a 4 deg. C sealed water calorimeter currently in operation at the University of Gent. This sealed water (SW) calorimeter is of the Domen type and has been improved in several aspects compared with its original design. The relevant correction factors for heat transport and for field perturbation are described. Using relative response measurements in ⁶⁰Co, we experimentally verified the relative heat defect for two distinct chemical systems, using two different detection vessel arrangements. The overall 1σ uncertainty on the absorbed dose to water at ⁶⁰Co based on this system amounts to 0.7%. (author)

[en] Full text: In recent years, international codes of practice based on absorbed dose to water standards have been published for the clinical reference dosimetry of external beams. It has become widely accepted that dosimetry of radiotherapeutic beams should be based on these standards. These codes of practice are a major improvement over earlier ones that used air kerma calibration factors as they are based on a calibration directly in a phantom in terms of the quantity of interest. The previous codes begin with calibration in air in terms of air kerma, then use theoretical and generic conversion factors to obtain dose to water that do not take account of chamber-to-chamber variation. Other good reasons for implementing the new codes are that they are conceptually simpler, include improved physical data and improve the consistency for various ionisation chamber types as well as between different beam types. TRS-3982,3 is a new Code of Practice (CoP) for reference dosimetry of external radiotherapy beams based on absorbed dose to, water calibrations and was published by the IAEA in a joint effort with the WHO, PAHO and ESTRO. It is the first CoP of its kind comprehensively covering all external radiotherapy beams except neutrons. The Radiotherapy Interest Group (RJG) of the ACPSEM has recommended that radiotherapy centres in Australia and New Zealand implement this CoP by the end of 2004. In this workshop, the general philosophy of the CoP will be outlined which will provide a framework for each of the individual subcodes. Although it represents just one of the potential implementations of the CoP, this workshop will deal only with dosimetry based on a cylindrical ionisation chamber with an absorbed dose calibration factor in 60Co from the standards laboratory. With the framework of the code in mind, it is straightforward to identify the basic steps that are required for measuring absorbed dose under reference conditions in a high-energy photon beam. The same is true for high-energy electron beams. However, the necessity of having an electron beam calibration factor for plane-parallel chambers and how to obtain it will be explained. Emphasis will be on practical aspects of implementing the CoP and the changes with respect to TRS-277 and TRS-381. Published experimental information on the expected differences to dosimetry based on the old codes will be discussed. Copyright (2004) Australasian College of Physical Scientists and Engineers in Medicine