Abstract Proton radiotherapy requires precise knowledge of the volumetric dose distribution. In proton beam delivery systems, based on narrow pencil beams, a contribution from small doses in low-intensity regions, consisting mainly of scattered protons, may have not negligible influence on total dose delivered to patient. Insufficient information about dose profile can cause underestimation of dose and potential delivery of inflated dose during hadrontherapy treatment. Presented work aims to verify applicability of diamond detectors, produced by Chemical Vapor Deposition method, for therapeutic proton beam profilometry at large fields. This requires the capability of measuring the core of the beam intensity profile (wide dynamic range) as well as its lateral spread (very high sensitivity) with a single device. INTRODUCTION Proton beam scanning (PBS) therapy is a type of particle therapy which uses a narrow, focused proton beam to irradiate precisely a well-defined treatment volume. The PBS delivers protons to each planned spot position step by step (using scanning magnets) and layer by layer (by changing the energy) until the entire cancer volume is treated. Based on a previously measured single spot parameters the Treatment Planning System calculates position, energy and ‘weight’ of each proton pencil beam. Proton beam model requires measurements of depth dose distributions of single beams for number of energies and measurements of spot profiles in air for a number of energies at different distances from isocenter. In this article we focused only on characteristic of lateral beam distributions. There are inconspicuous parts of proton beam profile, several centimeters away from the beam core which create the low-dose regions (tails). In therapeutic plans, consisting of large, overall number of spots, the dose contribution from low-intensity tails may have an impact, introducing discrepancy between delivered and prescribed dose. The spot profile characteristics are quite complex, because they depend on many parameters as, i.e. initial beam properties, transport conditions, beam focusing elements, construction of online beam monitors or presence of additional material in the nozzle, causing, i.e. Coulomb scattering and nuclear interactions(1). As a result, spot profile cannot be accurately modeled with single Gaussian dose distribution. It is necessary to precisely determine the realistic shape of the low-dose tails, also at large distances from the beam center(2). Moreover, as a tiny contribution from the spot tails influence the total dose, the precise beam profilometry at different therapy sites appears to be essential, because of their peculiarities. In Refs.(3, 4, 5) it have been described different methods for determination of beam spot characteristics. The group from Paul Scherrer Institute (PSI) measured primary spot characteristics in air and in water using a scintillator-CCD camera(6). The group at MD Anderson used both, gafchromic EBT film and the pin-point ionization chamber (2 mm diameter) to characterize spot profiles in air and in water(7, 8). The University of Pennsylvania(9) used a set of four films and high resolution Lynx® (scintillation screen/CCD camera) detector from IBA-Dosimetry (Schwarzenbruck, Germany) to determine 2D profiles of core and halo of spots at various positions in air and various depths for protons of various energies of PBS. The above-mentioned measurement methods demonstrated, however, some important limitations. The scintillation screen used at the PSI, have the dose sensitivity at the level of 0.006 Gy and saturates at ~1–2 Gy. Hence, the profile tails, with relative intensity below 0.3% with respect to central dose, are not directly measurable(6). The size of pin-point chamber used by MD Anderson group is comparable to the 3 mm sigma of the in-air pencil beams. This method is also time-consuming. The Lynx detector used at UPenn acquires data with 10-bit resolution (3 orders of magnitude of dynamic range), so the profile tails below 0.1% of central dose are not directly measurable without saturation of the beam core. In this work we present a verification of usefulness of diamond detectors, produced by Chemical Vapor Deposition method (CVD), for therapeutic proton beam profilometry and diagnostics. Proposed method, based on diamond detector, gives a results in very wide dynamic range, enables performing measurements in a single particle mode and has better background control than commercially available detectors. We attempt in this study to establish a novel measurement method using device based on diamond detector and employ the Lynx detector as an independent check. MATERIALS AND METHODS At the Cyclotron Center Bronowice (CCB) located in Krakow, Poland, we have the Proteus C-235 cyclotron produced by IBA (Belgium) accelerating protons to therapeutic energies. The CCB has been created specifically for cancer treatment (proton therapy), medical applications as well as scientific experiments. The Proteus is an isochronous cyclotron with compact, conventional magnet, delivering protons with energy of 230 MeV. This energy converts to a range of ~30 cm in water, what allows to reach all possible tumor locations in radiotherapy patient body. As the integral part the installation, two crucial devices: the energy degrader and energy selector, are responsible for providing monoenergetic proton beam with energies 70–230 MeV and currents up to 600 nA. At the CCB there are four facilities employing proton beams: eye treatment room and experimental hall, both with fixed, horizontal beam lines; two gantry treatment rooms equipped with dedicated scanning nozzles and rotating arms for isocentric irradiation of tumors (360° swing). Measurements of the single proton beam spot have been performed in a gantry room using scanning technique. To ensure a sufficient spatial resolution and dynamic range for profile measurements, single-crystal CVD diamond (scCVDD) detectors, with high charge collection efficiency, were used. The scCVDD detector was placed in front of the gantry nozzle, where the beam was passing through perpendicularly to the detector surface. Diamond as particle detector Diamond is being a promising material for ionizing radiation detection due to its unique properties as high radiation hardness, fast response due to the high carrier mobility and low noise (wide bandgap). Electric signal is generated in diamond detector when a high-energy charged particle goes through, causing an ionization along its track. The cloud of charge carriers, produced by impinging particle, travels in the presence of electric field, typically E = 1 V/μm. The high electric field causes a fast movement of electrons and the holes (causing a pulse signal) until they get collected by detector electrodes (contacts). The amplitude of the signal pulse increases with E and the pulse duration results shorter. The shorter the signal, the higher the pulse amplitude, because the total charge generated within detector volume (the area under signal pulse) is constant. Typical signal pulse width is <1 ns (FWHM) and after high amplification by low-noise electronics, allows to operate in a single particle mode (no pile-up), even for beam currents exceeding 40 pA in an area of 1 mm2. This conditions provide a potential for precise measurements of beam characteristics and its diagnostics in both time and space domains. Presented in this article results of measurements with 70 MeV proton beams have been performed with 100 μm thick scCVDD detector. Electronics Therapeutic protons with initial energies ranging from 70 to 230 MeV, deposit only a small fraction (~300 keV in the case of 70 MeV) of their primary energy in a thin ~100 μm scCVDD detector. This turns into a small charge generated ~3.5 fC by 70 MeV proton. As a result the most probable value of the pulse amplitude is ~160 μV at 50 Ohm input impedance and typical pulse duration ~1 ns for 5 mm2 for 100 μm diamond detector. Our CVDD detector structure was the electronic grade, single-crystal, produced by Element Six and operated at 200 V bias voltage. In order to detect such signals with good separation from the electronic noise, it is required to use wideband electronics (>1 GHz of bandwidth) with extremely low-noise level and high gain (>40 dB). This is especially important for easiness of discrimination of detected particles in particle counting applications. The scCVDD detectors provide intrinsically excellent S/N ratio, because of very low leakage current. As a result the most critical element essentially limiting the performance of detection systems is preamplifier. For this reason we decided to use a 2 GHz, low-noise (<19 μV input referred) PA-20 wideband preamplifier by INS Instruments with 20 dB signal gain, designed for low-noise applications(10). In order to increase the signal amplitude to assure the better separation from noise, we decided to connect two PA-20 preamplifiers in series. Very fast pulses generated by diamond detector, set also the requirement of using fast sampling devices. For our application we used the Teledyne LeCroy SDA 5000 oscilloscope with up to 20 GS/s sampling rate and 5 GHz of analog bandwidth. Detector signal of PBS on Gantry facility The Proteus C-235 cyclotron delivers a semi-continuous beam, which means that protons come in micropulses distant by ~9.4 ns. Under our measurement conditions, considering beam current density and the detector size, the probability of more than one proton in a micropulse is negligible. At lower beam currents, some micropulses remain empty (without particles). Figure 1 shows time distance between proton interaction events equal to ~75 ns, corresponding to seven empty micropulses. Figure 1. View largeDownload slide Exemplary signals by 70 MeV protons in 100 μm scCVD diamond detector, registered by the oscilloscope. Figure 1. View largeDownload slide Exemplary signals by 70 MeV protons in 100 μm scCVD diamond detector, registered by the oscilloscope. Determination of signal detection threshold In order to acquire accurate beam intensity profile by counting of single particle events, it is required to determine optimal amplitude threshold, between electronic noise and the signal amplitude. Data collected with the LeCroy SDA 5000, sampled at 10 GS/s with multiple signals in a single file (several files for each detector position), have been transferred to the PC for detailed, offline analysis. Prepared software searched for signal peaks in order to produce a distribution of signal amplitudes. In Figure 2, a histogram of signal waveform amplitudes, registered in the core position of the proton spot, is presented. The left distribution is due to the electronic, white noise with the mean value of ~0. The right histogram represents detected particle signals. Figure 2. View largeDownload slide Diamond pulses (registered signals) amplitude histogram. These signal waveforms have been acquired in the core position of the proton beam spot. Figure 2. View largeDownload slide Diamond pulses (registered signals) amplitude histogram. These signal waveforms have been acquired in the core position of the proton beam spot. The distribution is complemented with curves fitted to their specific regions. The Gaussian distribution is fitted to the electronic noise region, and the Landau function is fitted to proton signal events. Such an approach has allowed to determine an optimal threshold value of ~35 mV. The events with amplitudes above the given threshold can be considered as proton signals. This has been straightforward, mostly because of the preamplifiers performance, which allowed for good separation of signal pulses from the noise. RESULTS By counting in signals with amplitudes above the determined threshold within a measurement time window, one obtains a proton flux (intensity), for every detector position. The scan has been performed perpendicularly to the 70 MeV proton beam axis. In order to provide more accurate result, the measurement data have been normalized to a transversal monitor chamber, mounted in the gantry nozzle. This allowed to compensate for proton beam current momentary changes. Figure 3 presents a comparison of profile registered using two different techniques. The core region, from 1 to 100% intensity range, of the Gantry1_TPSinput profile was characterized with the commercial Lynx device. For the low-intensity regions (<1%) the TLD pellet detectors have been applied. Figure 3. View largeDownload slide Comparison of 70 MeV proton beam profiles, obtained with different detectors. The Gantry1_TPSinput profile (grey line) was obtained with the Lynx device (1–100%) and the TLD pellet detectors at low-intensity regions (<1%). The Gantry1_scCVDD profile (black dots) represents a profile measured with CVD diamond detector. For the intensity of 100% (core position) the CVD diamond count rate was ~2 × 106 particles/s and for 300 mm position (from the beam core) the count rate was ~5 particles/s. Figure 3. View largeDownload slide Comparison of 70 MeV proton beam profiles, obtained with different detectors. The Gantry1_TPSinput profile (grey line) was obtained with the Lynx device (1–100%) and the TLD pellet detectors at low-intensity regions (<1%). The Gantry1_scCVDD profile (black dots) represents a profile measured with CVD diamond detector. For the intensity of 100% (core position) the CVD diamond count rate was ~2 × 106 particles/s and for 300 mm position (from the beam core) the count rate was ~5 particles/s. To get the full profile, the Gantry1_TPSinput had to be merged, filtered and interpolated, to achieve a desired shape and good resolution in entire intensity range. Instead, the Gantry1_scCVDD profile represents a proton flux spatial distribution, measured by proton counting technique with diamond detector. The core region of the profile at gantry is in a good agreement with the data measured using the Lynx detector, however, in a region with relative intensities <0.01%, the difference becomes not negligible. Moreover, this is increasing with the distance from the core of the beam, where the high system sensitivity is strongly required. It is a consequence of the fact that the proposed method, with diamond detector can be considered as ‘noiseless’. While it operates in single particle (counting) mode, the TLD pellets show dose sensitivity above microgray levels(11). It is noticeable, that low-dose regions should also be taken into account during TPS configuration of beam characteristics. Exploiting diamond detectors operating in a single particle regime, allows to reach more than 6 orders of magnitude of dynamic range, hence ~2 × 106 protons/s in the beam core and ~few protons/s at the beam profile tails. CONCLUSIONS The intensity profile of the proton beam delivered by the Proteus C-235 cyclotron has been characterized with different methods and detectors. The obtained results demonstrate, that proton counting technique with scCVD diamonds offers significant advantages with respect to other detectors. Providing a capability of measuring beam intensities with a dynamic range exceeding 106, while preserving single particle sensitivity, is crucial for both, the beam core and low-intensity regions (tails). Within the presented work we have developed as well a robust data analysis framework and experimental procedure, applicable to therapeutic proton beams of different characteristics. Accurate and precise information about proton beam intensity distribution is of great importance in hadrontherapy planning, in order to avoid underestimation of the dose during treatment. The proposed method of determination of spatial intensity distribution ensures as well the required spatial resolution as high dynamics using single detector. The results of this study may provide helpful information in order to improve accuracy of dose calculation. The proposed technique can improve beam profile characterization, especially important for spot scanning particle therapy. It is feasible to achieve small pixel size (<0.5 mm) which can satisfy spatial resolution requirements. Providing the capability to measure intensity with very wide dynamic range, exceeding 106, it makes possible to obtain valuable and specific beam information within entire intensity range. Operation of the diamond detector in single particle regime, can also be designed as an online system, replacing sampling device as oscilloscope in order to avoid time-consuming data transfer and offline analyses. REFERENCES 1 Gottschalk , B. et al. . Nuclear halo of a 177 MeV proton beam in water: theory, measurement and parameterization . Med. Phys. arXiv:1409.1938 (2014). 2 Li , Y. et al. . Beyond Gaussians: a study of single-spot modeling for scanning proton dose calculation . Phys. Med. Biol. 57 , 983 – 997 ( 2012 ). Google Scholar CrossRef Search ADS PubMed 3 Pedroni , E. et al. . Experimental characterization and physical modelling of the dose distribution of scanned proton pencil beams . Phys. Med. Biol. 50 , 541 – 561 ( 2005 ). Google Scholar CrossRef Search ADS PubMed 4 Gillin , M. T. et al. . Commissioning of the discrete spot scanning proton beam delivery system at the University of Texas M.D. Anderson Cancer Center, Proton Therapy Center, Houston . Med. Phys. 37 , 154 – 163 ( 2009 ). Google Scholar CrossRef Search ADS 5 Clasie , B. et al. . Golden beam data for proton pencil-beam scanning . Phys. Med. Biol. 57 , 1147 – 1158 ( 2012 ). Google Scholar CrossRef Search ADS PubMed 6 Boon , S. N. et al. . Performance of a fluorescent screen and CCD camera as a two-dimensional dosimetry system for dynamic treatment techniques . Med. Phys. 27 , 2198 – 2208 ( 2000 ). Google Scholar CrossRef Search ADS PubMed 7 Soukup , M. et al. . A pencil beam algorithm for intensity modulated proton therapy derived from Monte Carlo simulations . Phys. Med. Biol. 50 , 5089 – 5104 ( 2005 ). Google Scholar CrossRef Search ADS PubMed 8 Zhu , X. R. et al. . Commissioning dose computation models for spot scanning proton beams in water for a commercially available treatment planning system . Med. Phys. 40 , 041723 ( 2013 ). Google Scholar CrossRef Search ADS PubMed 9 Lin , L. et al. . A novel technique for measuring the low-dose envelope of pencil-beam scanning spot profiles . Phys. Med. Biol. 58 , N171 – N180 ( 2013 ). Google Scholar CrossRef Search ADS PubMed 10 INS Instruments . Wideband Amplifiers. [Online: 14-07-2017] http://widebandamplifiers.com. 11 Obryk , B. et al. . Method of thermoluminescent measurement of radiation doses from micrograys up to a megagray with a single LiF:Mg,Cu,P detector . Radiat. Prot. Dosim. 144 , 543 – 547 ( 2011 ). Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: email@example.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
Radiation Protection Dosimetry – Oxford University Press
Published: Aug 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
All the latest content is available, no embargo periods.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud