Analysis of the CONRAD computational problems expressing only stochastic uncertainties: neutrons and protons.

Within the scope of CONRAD (A Coordinated Action for Radiation Dosimetry) Work Package 4 on Computational Dosimetry jointly collaborated with the other research actions on internal dosimetry, complex mixed radiation fields at workplaces and medical staff dosimetry. Besides these collaborative actions, WP4 promoted an international comparison on eight problems with their associated experimental data. A first set of three problems, the results of which are herewith summarised, dealt only with the expression of the stochastic uncertainties of the results: the analysis of the response function of a proton recoil telescope detector, the study of a Bonner sphere neutron spectrometer and the analysis of the neutron spectrum and dosimetric quantity H(p)(10) in a thermal neutron facility operated by IRSN Cadarache (the SIGMA facility). A second paper will summarise the results of the other five problems which dealt with the full uncertainty budget estimate. A third paper will present the results of a comparison on in vivo measurements of the (241)Am bone-seeker nuclide distributed in the knee. All the detailed papers will be presented in the WP4 Final Workshop Proceedings.

W values of protons in liquid water.

The W values of protons in liquid water were calculated for energies from 0.1 keV to 10 MeV using the continuous slowing down approximation as well as three models for the calculation of the differential ionisation cross-sections of water for protons published in recent years. The W values determined by means of the three models differ only marginally from each other and lie between 25 and 26 eV at proton energies >5 MeV. This high-energy W value is approximately 3 eV lower than that in water vapour.

New descriptors of radiation quality based on nanodosimetry, a first approach.

After a short overview on the latest developments in nanodosimetry, measured frequency distributions of ionisation cluster size caused by 4.6 MeV alpha-particles or low-energy electrons in 'nanometric' volumes of nitrogen are compared with cluster-size distributions for liquid water cylinders that are equal in size to segments of DNA of 10 base-pairs length. Such frequency distributions are, to a greater part, governed by the same basic physical interaction data as those to be expected, if charged particles interact with DNA segments. Quantities derived from ionisation cluster-size distributions should, therefore, behave as a function of radiation quality similarly to the yields of single or double strand breaks in the DNA. To test this assumption, extensive Monte Carlo simulations were performed for electrons in the energy range between 12.5 eV and 100 keV for protons at energies between 0.7 MeV and 250 MeV and for alpha-particles in the energy range between 2 MeV and 100 MeV. The results are then compared with the yields of single- or double-strand breaks in the DNA, taken from the literature.

Proton-induced frequency distributions of ionization cluster size in propane.

The frequency distribution of clustered ionizations produced by a proton beam was measured in a nanodosimetric volume of the size of a DNA segment by means of an ion-counting nanodosimeter in the energy range from 0.4 to 3.5 MeV. In order to meet the needs of the ion-counting nanodosimeter, the accelerator's primary beam was reduced in intensity by means of Rutherford scattering. The comparison between experimental results and Monte Carlo simulations show a good agreement in the energy dependence of the mean cluster size, while the experimental cluster size distributions show a higher amount of large ionization clusters compared with those obtained with the simulations.

Measurement of the x-ray mass energy-absorption coefficient of air using 3 keV to 10 keV synchrotron radiation.

For the first time absolute photon mass energy-absorption coefficients of air in the energy range 3 keV to 10 keV have been measured with relative standard uncertainties less than 1%, significantly smaller than those of up to 5% assumed hitherto for calculated data. Monochromatized synchrotron radiation was used to measure both the total radiant energy by means of silicon photodiodes calibrated against a cryogenic radiometer and the fraction of radiant energy that is deposited in dry air by means of a free air ionization chamber. The measured ionization charge was converted into energy absorbed in air by calculated effective W values of photons as a function of their energy based on new measurements of the W values in dry air for electron kinetic energies between 1 keV and 7 keV, also presented in this work. The measured absorption coefficients were compared with state-of-the art calculations and found to agree within 0.7% with data calculated earlier by Hubbell at energies above 4 keV but were found to differ by values up to 2.1% at 10 keV from more recent calculations of Seltzer.

Ionisation cluster-size formation by electrons: from macroscopic to nanometric target sizes.

An indispensable prerequisite for a deeper understanding of specified physical, chemical or biological changes initiated in matter when being exposed to ionising radiation is a detailed knowledge of particle track structure. Here, the structure of electron tracks is of particular interest since electrons are set in motion in large numbers as secondary particles during the slow down of any kind of ionising radiation in matter. From the point of view of radiation induced early damage to genes and cells, which starts with the early damage to segments of the DNA molecule, the most effective secondary electrons are those at energies of a few hundred eV since the yield of double-strand breaks induced by such electrons in the DNA shows a maximum. This can be explained by the fact that in water cylinders, 2 nm in diameter and height (as a substitute to small segments of the DNA), the probability of the electron-induced formation of ionisation cluster sizes greater than or equal to two is highest also at initial electron energies of a few hundred eV. In view of this promising feature of ionisation cluster-size distributions formed by low-energy electrons in nanometric targets of liquid water for explaining particular radio-biological endpoints, it is the aim of the present work to investigate the properties of cluster-size formation by electrons as a function of target size. Here, main emphasis is laid on the behaviour of cluster-size distributions if the target size is reduced from macroscopic to nanometric volumes.

Nanodosimetry, the metrological tool for connecting radiation physics with radiation biology.

It is generally accepted that the early damage to genes or cells by ionising radiation starts with the early damage to segments of the DNA, at least, to the greater part. This damage is the result of the spatial distribution of inelastic interactions of single ionising particles within the DNA or in its neighbourhood and is, in consequence, determined by the stochastics of particle interactions in volumes a few nanometre in size. Due to the latter fact radiation damage strongly depends on radiation quality and cannot be described satisfactorily in detail by macroscopic quantities like absorbed dose or linear energy transfer (LET). It can, however, be described approximately by the probability distribution of ionisation cluster-size formation in nanometric target volumes of liquid water (nanodosimetry). One aim of nanodosimetry is, therefore, to measure the radiation induced frequency distribution of ionisation cluster-size formation in liquid water, as a substitute for sub-cellular material, in volumes which are comparable in size with those of the most probable radio-sensitive volumes of biological systems. After a short description of the main aspects of cluster-size formation by ionising particles, an overview is given about the measuring principles applied in present-day nanodosimetric measurements. Afterwards, physical principles are discussed which can be used to convert ionisation cluster-size distributions measured in gases into those caused by ionising radiation in liquid water. In a final section, the probability distribution of ionisation cluster-size formation in liquid water is discussed from the point of view of damage formation to segments of the DNA.

Formation of ion clusters by low-energy electrons in nanometric targets: experiment and Monte Carlo simulation.

The first experimental data on the distribution of ionisation cluster size produced by low energy electrons in a target cylinder of nitrogen, 3.5 nm in diameter at unit density, which is equivalent to approximately 2 nm in liquid water are presented. In the experiment, nanometric targets were simulated in the so-called Jet Counter. It consists of a pulse-operated valve which injects an expanding jet of nitrogen into an interaction chamber where a cylindrical sensitive volume is created. This sensitive volume was irradiated by electrons at 300 eV, 500 eV and 1 keV, emitted by an electron gun. The distribution of ionisation cluster size was measured using the single-ion-counting method and compared with the results of Monte Carlo simulation.

Ionization ranges of protons in water vapour in the energy range 1-100 keV.

Projected ranges of protons in water vapour were experimentally determined for proton energies from 1 to 100 keV by counting the total number of ionizations produced by protons during their slow down. Using these experimental ranges and semiemprical detour factors, the stopping powers of water vapour for protons were derived and compared with semiempirical data.

Simulation of the measured ionisation-cluster distributions of alpha-particles in nanometric volumes of propane.

In the last years, the probability of the formation of ionisation clusters by primary alpha particles at 5.4 MeV in nanometric volumes of propane (20.6 and 24.0 nm in a material of density 1.0 g cm(-3)) was studied experimentally and by Monte Carlo simulation. Calculations were performed taking into account the single electron detection efficiency of the track-nanodosimetric counter, which was estimated on the base of Monte Carlo calculations of electron transport inside the detector. Now a new evaluation of the efficiency has been performed, pointing out a value lower than previously estimated. Besides, the efficiency of the counter in resolving temporally the collected electrons has been calculated, together with its effect on the measured distribution. On the base of these evaluations, a new comparison has been performed between measurements and calculations, pointing out a better agreement than previously reported.

Bayesian reconstruction of nanodosimetric cluster distributions at 100% detection efficiency.

Ionisation spectra in nanometric volumes at a given distance from a charged particle track are obtained by using electron (or ion) gas detectors, having non-uniformly distributed detection efficiency. Therefore, such spectra should be properly processed in order to reconstruct the frequency distribution of clusters really produced in the detector gas. A Bayesian unfolding has been applied to ionisation distributions due to 5.4 MeV alpha particles in a 20-nm site obtained by Monte Carlo simulations, taking into account different detection efficiency conditions. It will be shown that Bayesian analysis provides a valid tool for reconstructing the true ionisation distributions, well beyond the maximum measured cluster size.

Dependence of nanodosimetric spectra on the sensitive volume length and ion drift in an ion-counting nanodosemeter.

Nanodosimetric spectra, measured in a well-defined ionisation sensitive volume of an ion-counting gaseous nanodosemeter, may have a valuable predictive value of radiation damage to DNA. In such devices, the distributions of radiation-induced ions are measured after their drift in gas. The sensitive-volume size, corresponding to a DNA segment length, can be tuned by selecting an appropriate time window for ion counting; the method's accuracy depends on the velocity distribution of the drifting ions. The results of ion-drift measurements in an ion-counting nanodosemeter were used for the precise calculation of its sensitive volume length. Monte Carlo simulations of nanodosimetric spectra, performed with the obtained data, are in good agreement with experimental data. The method's limitations, arising from the spread of drift velocities, are discussed.

First attempts at prediction of DNA strand-break yields using nanodosimetric data.

We present the first results of our attempts to correlate yields of ionisation clusters in a gas model of DNA and corresponding double-strand break (DSB) yields in irradiated plasmids, using a simple statistical model of DNA lesion formation. Based on the same statistical model, we also provide a comparison of simulated nanodosimetric data for electrons and published DSB yields obtained with the PARTRAC code.

Ion-counting nanodosemeter with particle tracking capabilities.

An ion-counting nanodosemeter (ND) yielding the distribution of radiation-induced ions in a low-pressure gas within a millimetric, wall-less sensitive volume (SV) was equipped with a silicon microstrip telescope that tracks the primary particles, allowing correlation of nanodosimetric data with particle position relative to the SV. The performance of this tracking ND was tested with a broad 250 MeV proton beam at Loma Linda University Medical Center. The high-resolution tracking capability made it possible to map the ion registration efficiency distribution within the SV, for which only calculated data were available before. It was shown that tracking information combined with nanodosimetric data can map the ionisation pattern of track segments within 150 nm-equivalent long SVs with a longitudinal resolution of approximately 5 tissue-equivalent nanometers. Data acquired in this work were compared with results of Monte Carlo track structure simulations. The good agreement between 'tracking nanodosimetry' data acquired with the new system and simulated data supports the application of ion-counting nanodosimetry in experimental track-structure studies.

Pitfalls and modelling inconsistencies in computational radiation dosimetry: lessons learnt from the QUADOS intercomparison. Part II: Photons, electrons and protons.

'QUADOS', a concerted action of the European Commission, has promoted an intercomparison aimed at evaluating the use of computational codes for dosimetry in radiation protection and medical physics. This intercomparison was open to all users of radiation transport codes. Eight problems were selected for their relevance to the radiation dosimetry community, five of which involved photon and proton transport. This paper focuses on a discussion of lessons learned from the participation in solving the photon and charged particle problems. The lessons learned from the participation in solving the neutron problems are presented in a companion paper (in this issue).

Pitfalls and modelling inconsistencies in computational radiation dosimetry: lessons learnt from the QUADOS intercomparison. Part I: Neutrons and uncertainties.

The QUADOS EU cost shared action conducted an intercomparison on the usage of numerical methods in radiation protection and dosimetry. The eight problems proposed were intended to test the usage of Monte Carlo and deterministic methods by assessing the accuracy with which the codes are applied and also the methods used to evaluate uncertainty in the answer gained through these methods. The overall objective was to spread good practice through the community and give users information on how to assess the uncertainties associated with their calculated results.

Evaluation of lesion clustering in irradiated plasmid DNA.

PURPOSE: To measure the yield of DNA strand breaks and clustered lesions in plasmid DNA irradiated with protons, helium nuclei, and y-rays. MATERIALS AND METHODS: Plasmid DNA was irradiated with 1.03, 19.3 and 249 MeV protons (linear energy transfer = 25.5, 2.7, and 0.39 keV microm(-1) respectively), 26 MeV helium nuclei (25.5 keV microm) and gamma-rays (137Cs or 60Co) in phosphate buffer containing 2 mM or 200 mM glycerol. Single-and double-strand breaks (SSB and DSB) were measured by gel electrophoresis, and clustered lesions containing base lesions were quantified by converting them into irreparable DSB in transformed bacteria. RESULTS: For protons, SSB yield decreased with increasing LET (linear energy transfer). The yield of DSB and all clustered lesions seemed to reach a minimum around 3 keV microm(-1). There was a higher yield of SSB, DSB and total clustered lesions for protons compared to helium nuclei at 25.5 keV microm(-1). A difference in the yields between 137Cs and 60Co gamma-rays was also observed, especially for SSB. CONCLUSION: In this work we have demonstrated the complex LET dependence of clustered-lesion yields, governed by interplay of the radical recombination and change in track structure. As expected, there was also a significant difference in clustered lesion yields between various radiation fields, having the same or similar LET values, but differing in nanometric track structure.

Nanodosimetry, from radiation physics to radiation biology.

In view of the fact that early damage to genes and cells by ionising radiation starts with the early damage to segments of the DNA, it is a great challenge to radiation research to describe the general behaviour of ionising radiation in nanometric target volumes (nanodosimetry). After summarising basic aspects of nanodosimetry, an overview is given about its present state. As far as experimental procedures are concerned, main emphasis is laid on single-ion counting and single-electron counting methods, which use millimetric target volumes filled with a low-pressure gas to simulate nanometric target volumes at unit density. Afterwards, physical principles are discussed, which can be used to convert experimental ionisation cluster-size distributions into those caused by ionising radiation in liquid water. In the final section, possibilities are analysed of how to relate parameters derived from the probability of cluster-size formation in liquid water to parameters derived from radiobiological experiments.

QUADOS intercomparison: a summary of photon and charged particle problems.

QUADOS, a Concerted Action of the European Commission, has promoted an intercomparison aimed at evaluating the use of computational codes for dosimetry in radiation protection and medical physics. This intercomparison was open to all users of radiation transport codes. Eight problems were selected for their relevance to the radiation dosimetry community, five of which involved photon and proton transport. This paper focuses on the analysis of the photon and charged particle problems. The neutron problems were presented in a paper at the NEUDOS9 conference.

Nanodosimetric track structure in homogeneous extended beams.

Physical aspects of particle track structure are important in determining the induction of clustered damage in relevant subcellular structures like the DNA and higher-order genomic structures. The direct measurement of track-structure properties of ionising radiation is feasible today by counting the number of ionisations produced inside a small gas volume. In particular, the so-called track-nanodosimeter, installed at the TANDEM-ALPI accelerator complex of LNL, measures ionisation cluster-size distributions in a simulated subcellular structure of dimensions 20 nm, corresponding approximately to the diameter of the chromatin fibre. The target volume is irradiated by pencil beams of primary particles passing at specified impact parameter. To directly relate these measured track-structure data to radiobiological measurements performed in broad homogeneous particle beams, these data can be integrated over the impact parameter. This procedure was successfully applied to 240 MeV carbon ions and compared with Monte Carlo simulations for extended fields.