Abstract Gafchromic EBT3 films are applied in proton radiotherapy for 2D dose mapping because they demonstrate spatial resolution well below 1 mm. However, the film response must be corrected in order to reach the accuracy of dose measurements required for the clinical use. The in-house developed AnalyseGafchromic software allows to analyze and correct the measured response using triple channel dose calibration, statistical scan-to-scan fluctuations as well as experimentally determined dose and LET dependence. Finally, the optimized protocol for evaluation of response of Gafchromic EBT3 films was applied to determine 30 × 40 cm2 dose profiles of the scanning therapy unit at the Cyclotron Centre Bronowice, CCB in Krakow, Poland. INTRODUCTION In modern proton therapy facilities, the pencil scanning beam (PBS) irradiation technique is now becoming broadly used for patient treatment. PBS usually leads to improved dose distributions but requires careful quality assurance (QA). Dosimetric verification of PBS treatment plans requires 2D dosimeters due to the possible presence of hot and cold spots. Currently, the MatriXXPT (IBA Dosimetry) array of ionization chambers is applied at CCB to simultaneously measure dose and evaluate the spatial dose distribution in PBS radiotherapy(1). However, the MatriXX spatial resolution is limited due to the size of the chambers and the distance between them (4.5 and 7.6 mm, respectively). As an alternative, Gafchromic (Ashland Inc.) EBT3 self-developing films with spatial resolution below 1 mm can be applied for QA(2, 3). Many studies(4–7) have shown the critical impact of beam energy distribution (simplified as dependence on linear energy transfer (LET)) on detector efficiency of the Gafchromic films. The film can be digitized by scanning it with a flatbed color scanner. Therefore, in addition to EBT3 film calibration, the scanner uniformity and stability corrections have to be applied to increase the accuracy of the system. The aim of this work was to develop a dosimetric system, composed of the EBT3 film, a scanner and a set of correction algorithms to recalculate the measured response into absorbed dose in water for clinical proton beams. Finally, an advanced home-developed software tool AnalyseGafchromic was developed to facilitate the usage of the films in clinical conditions. MATERIAL AND METHODS Irradiations and detectors readout All irradiations have been performed with PBS using dedicated scanning nozzle at the Cyclotron Centre Bronowice CCB (Krakow) produced by Ion Beam Application (IBA). The Proteus C-235 cyclotron with energy selector delivers proton beam in energy range between 70 and 226 MeV. For lower energies, the range shifter is installed at the nozzle beam exit. Gafchromic EBT3 8 × 10 (lot #03 171 401) films and, for scanner uniformity measurement, Gafchromic EBT3 1417 (lot #12 171 302) were used. Detector readouts were performed 24 h after irradiations using the Epson Expression 11 000XL flatbed scanner in 48-bit (16-bit per color channel) transmission mode with the resolution of 75 dpi. Each piece of film has been positioned along the central axis of the scanner in landscape orientation, as recommended by the manufacturer. Data were analyzed using our in-house developed tool AnalyseGafchromic, which is created in the MATLAB 2016a software environment. Triple channel calibration Dose calibration curves were obtained from readout of seven film pieces (4 × 4 cm2) irradiated with proton beam at doses of 0.30, 0.59, 0.99, 1.98, 3.95, 7.91 and 11.86 Gy at the depth of 7 cm in the solid water phantom. The irradiation field had a 10 × 10 cm2 rectangular shape, a range of 10 g/cm2 and a modulation of 5 g/cm2. All films were positioned along the long axis of the scanner and scanned all together in a single scan with an unexposed (background) film. For each detector a ROI of 2 × 2 cm2 was analyzed to determine mean pixel value vs. delivered dose. Calibration was performed for each color channel k separately. Pixel values (PV) have been normalized to 216-1 (maximum unsigned integer value of 16-bit scale). Calibration function was fitted independently for each channel k by the following function: Dk(nPVk)=(fk*nPVk+gk)/(nPVk+hk) (1) where, fk, gk and hk are the fitting parameters, nPVk is the normalized pixel value for color channel k and Dk is the dose value calculated based on channel k. Calibration function for each channel can be used for dose calculation separately, but in practice the red channel is usually used. In this work the dosimetry accuracy was enhanced by combining response of all three channels(8, 9). This triple channel dosimetry algorithm was implemented in our MATLAB software. Scan-to-scan correction In order to minimize the impact of the scanner sensitivity fluctuations, exposed films have been always scanned together with two calibration EBT3 films: one with the lowest (0.00 Gy) and another with the highest (11.86 Gy) calibrated dose. According to the procedure introduced by Lewis et al.(10) the ratio of pixel values of the calibration films obtained during the calibration procedure and during each particular scan allows to correct calibration functions for scanner fluctuations. In order to determine the scanning reproducibility, a film irradiated with a dose of 2 Gy had been scanned together with two reference calibration films (0.00 and 11.86 Gy) 30 times with the time gap of 1 min between scans. Then the dose for each scan has been determined in 2 × 2 cm2 ROI both with and without calibration function modifying. Scanner non-uniformity correction The lateral scanner artefact (or parabola effect) is an effect rooted in the scanner architecture, which causes an non-uniform scanner response along the axis parallel to the elongated light source (short scanner axis). Poppinga et al.(11) proposed a method of parabola effect correction by fitting parabola curves to scanner profiles in the red channel. In this work the method was applied to all three RGB channels in order to apply the scanner non-uniformity correction together with individual channel calibration and scan-to-scan correction. Nine pieces of films 5 × 35 cm2 were irradiated with uniform doses between 0.2 and 17.5 Gy and one was left unexposed. During readout, films were positioned in the direction of the elongated light source. Each film was scanned nine times—at the different position along the long scanner axis. ROIs of 2.6 × 30.9 cm2 were analyzed. As the homogeneity of EBT3 film sheets has been found to be better than 1.5%(12), no additional correction for film uniformity has been applied. The final profile of each film (for each R, G, B channel separately), was an average for all profiles acquired for this detector, fitted with parabolic function(11): PVk(x)=ck+ak(x−bk)2 (2) where, PVk is pixel value in channel k for lateral position x, bk is the x coordinate of parabola vertex, ck is the pixel value for x = bk. ak is the parameter which describes the parabolas curvature. In order to evaluate the uniformity correction method, one Gafchromic film was exposed to a uniform 30 × 40 cm2 proton field at 150 MeV and a surface dose of ~2 Gy. The film was then subsequently analyzed with the scanner uniformity correction enabled and disabled. Proton LET correction The response of Gafchromic films is known to decrease with increasing ionization density quantified e.g. by unrestricted Linear Energy Transfer, LET(4–7). Films were irradiated sandwiched in a RW3 solid water phantom using PBS at CCB, Krakow. A set of 13 Spread Out Bragg Peaks (SOBP) were prepared, with modulations ranging from 5 to 15 cm. The film was always irradiated in the middle of the SOBP. The proton LET correction was determined by normalizing the measured detector response to dose in phantom determined with Markus ionization chamber. The track-averaged LETT, corresponding to the center of SOBP, was simulated using SHIELD-HIT12A (v.0.6.0) proton transport code. The LETT was calculated including all secondary particles produced by nuclear reactions. Finally the dose is calculated with the triple channel calibration method by using the scan-to-scan modified calibration functions and corrected by scanner non-uniformity and the LET effectiveness. RESULTS AND DISCUSSION The calibration functions Dk(nPVk) are shown in Figure 1. Fitting parameters of calibration function Dk(nPV) are presented in Table 1 together with their uncertainties. Table 1. Fit coefficients obtained for calibration functions. 95% Confidence intervals are used for uncertainties. Channel k fk gk hk R2 Red (R) −3.886(74) 2.378(37) −0.0209(23) 0.99999 Green (G) −7.30(24) 4.48(14) −0.0014(72) 0.99998 Blue (B) −13.8(13) 5.14(47) −0.062(15) 0.99991 Channel k fk gk hk R2 Red (R) −3.886(74) 2.378(37) −0.0209(23) 0.99999 Green (G) −7.30(24) 4.48(14) −0.0014(72) 0.99998 Blue (B) −13.8(13) 5.14(47) −0.062(15) 0.99991 Table 1. Fit coefficients obtained for calibration functions. 95% Confidence intervals are used for uncertainties. Channel k fk gk hk R2 Red (R) −3.886(74) 2.378(37) −0.0209(23) 0.99999 Green (G) −7.30(24) 4.48(14) −0.0014(72) 0.99998 Blue (B) −13.8(13) 5.14(47) −0.062(15) 0.99991 Channel k fk gk hk R2 Red (R) −3.886(74) 2.378(37) −0.0209(23) 0.99999 Green (G) −7.30(24) 4.48(14) −0.0014(72) 0.99998 Blue (B) −13.8(13) 5.14(47) −0.062(15) 0.99991 Figure 1. View largeDownload slide Calibration curves used for calculation of dose from the normalized pixel value of the given channel, nPVk. The coefficients to calculate the shown calibration curves according to Equation 1 are given in Table 1. Figure 1. View largeDownload slide Calibration curves used for calculation of dose from the normalized pixel value of the given channel, nPVk. The coefficients to calculate the shown calibration curves according to Equation 1 are given in Table 1. Films irradiated with 2 Gy were scanned 30 times and the relative standard deviation of the response was calculated to both the corrected and uncorrected scans. The scan-to-scan corrections reduced the relative standard deviation of the measured film response from 0.93 to 0.39% Figure 2 shows raw scanner lateral profiles in red channel measured for films irradiated with doses from 0 to 17.5 Gy. It is has been noticed that the non-uniformity increases with dose. As the ak from Equation 2 denotes the parabola curvature and ck refers to dose, the scanner non-uniformity dose dependence can be quantified by ak(ck)(11). Figure 2. View largeDownload slide Raw scanner lateral profiles in the red channel, PVR (x), measured for films irradiated with different doses. Figure 2. View largeDownload slide Raw scanner lateral profiles in the red channel, PVR (x), measured for films irradiated with different doses. The dose effect is the most significant for the red channel, where the highest ck values were determined. Moreover, for this channel ak(ck) dependence is monotonic and may be quantified by fitting parabolic function, a modified version of formula from Poppinga et al.(11) aR(cR)=rcR2+qcR+p (3) Equation 2, for the red channel, can be written as Equation (4) or (5). PVR(x)=cR+(rcR2+qcR+p)(x−bR)2 (4) PVR(x)=r(x−bR)2cR2+(1+q(x−bR)2)cR+p(x−bR)2 (5)cR value is the parabola vertex (maximum) and it represents the pixel value in the absence of the parabola effect. Equation 5 can be solved analytically to derive cR based on fitting parameters p, q, r and bR. bR was found to be independent of dose and is constant along the central axis of the scanner. For the green and blue channels, the parabolic interpolation of PVk(x) for different dose levels was not relevant, and the dose response is calculated as the mean profile for all doses. The relative efficiency, η, defined as a ratio of doses measured by Gafchromic and IC, plotted versus track-averaged LETT is shown in Figure 3. For SOBP with modulation of 10 cm η typically does not exceed 1%. Only at the lowest modulation depth close to the distal edge, the particle spectrum contains more stopping protons. Then, the correction for η is significant and should be taken into account. Figure 3. View largeDownload slide The relative efficiency of EBT3 films as a function of track-averaged LETT of proton beams in a solid water phantom. The total uncertainty of film effectiveness consists of the uncertainty of absolute dose determination with Markus chamber and the film readouts reproducibility. Figure 3. View largeDownload slide The relative efficiency of EBT3 films as a function of track-averaged LETT of proton beams in a solid water phantom. The total uncertainty of film effectiveness consists of the uncertainty of absolute dose determination with Markus chamber and the film readouts reproducibility. Finally, the non-uniformity correction was applied to films irradiated with a 30 × 40 cm2 monoenergetic proton field of 150 MeV (Figure 4). The dose uniformity in the plateau region is within 3% on the corrected image compared with up to 15% difference when the correction is disabled. Figure 4. View largeDownload slide Relative dose of 30 × 40 cm2 measured with Gafchromic films in monoenergetic proton field of energy 150 MeV with scanner non-uniformity enabled and disabled. Figure 4. View largeDownload slide Relative dose of 30 × 40 cm2 measured with Gafchromic films in monoenergetic proton field of energy 150 MeV with scanner non-uniformity enabled and disabled. All the calibration and correction coefficients, developed within this work was built into the dedicated software tool AnalyseGafchromic and is routinely applied for QA measurements at the CCB proton therapy units. CONCLUSIONS The paper presents a method of determining a set of calibration and correction factors needed to modify the measured response of Gafchromic film for dosimetry of scanning proton beams. The Poppinga et al.(11) correction algorithm for scanner uniformity was modified to be used for all three red (R), green (G) and blue (B) channels simultaneously. However, only for the R channel monotonic dose dependence of scanner response was observed. For G and B dose dependence was not relevant. Scanner fluctuation were reduced by normalizing the measured signal from calibration films. LET corrections were found necessary only for measurements in the vicinity of distal fall-off of the SOBP. Tests performed for 30 × 40 cm2 uniform proton fields show that using the implemented algorithm, the deviation of the measured dose was reduced from ~15 to 3%. Finally, all corrections were incorporated to the in-house developed AnalyseGafchromic software which is now applied for the routine QA measurements at CCB. REFERENCES 1 Lin , L. , Kang , M. , Solberg , T. D. , Mertens , T. , Baumer , C. , Ainsley , C. G. and McDonough , J. E. Use of a novel two‐dimensional ionization chamber array for pencil beam scanning proton therapy beam quality assurance . J. Appl. Clin. Med. Phys. 16 ( 3 ), 270 – 276 ( 2015 ). Google Scholar CrossRef Search ADS 2 Butson , M. J. , Peter , K. , Cheung , T. and Metcalfe , P. Radiochromic film for medical radiation dosimetry . Mater. Sci. Eng. R. Rep. 41 ( 3 ), 61 – 120 ( 2003 ). Google Scholar CrossRef Search ADS 3 Devic , S. Radiochromic film dosimetry: past, present, and future . Phys. Med. 27 ( 3 ), 122 – 134 ( 2011 ). Google Scholar CrossRef Search ADS PubMed 4 Zhao , L. and Das , I. J. Gafchromic EBT film dosimetry in proton beams . Phys. Med. Biol. 55 ( 10 ), N291 ( 2010 ). Google Scholar CrossRef Search ADS PubMed 5 Gambarini , G. , Regazzoni , V. , Grisotto , S. , Artuso , E. , Giove , D. , Borroni , M. , Carrara , M. , Pignoli , E. , Mirandola , A. and Ciocca , M. Measurements of spatial distribution of absorbed dose in proton therapy with Gafchromic EBT3 ( 2014 ). 6 Kirby , D. , Green , S. , Palmans , H. , Hugtenburg , R. , Wojnecki , C. and Parker , D. LET dependence of GafChromic films and an ion chamber in low-energy proton dosimetry . Phys. Med. Biol. 55 ( 2 ), 417 ( 2009 ). Google Scholar CrossRef Search ADS PubMed 7 Martišíková , M. and Jäkel , O. Dosimetric properties of Gafchromic® EBT films in monoenergetic medical ion beams . Phys. Med. Biol. 55 ( 13 ), 3741 ( 2010 ). Google Scholar CrossRef Search ADS PubMed 8 Micke , A. , Lewis , D. F. and Yu , X. Multichannel film dosimetry with nonuniformity correction . Med. Phys. 38 ( 5 ), 2523 – 2534 ( 2011 ). Google Scholar CrossRef Search ADS PubMed 9 Mayer , R. R. , Ma , F. , Chen , Y. , Miller , R. I. , Belard , A. , McDonough , J. and O’Connell , J. J. Enhanced dosimetry procedures and assessment for EBT2 radiochromic film . Med. Phys. 39 ( 4 ), 2147 – 2155 ( 2012 ). Google Scholar CrossRef Search ADS PubMed 10 Lewis , D. , Micke , A. , Yu , X. and Chan , M. F. An efficient protocol for radiochromic film dosimetry combining calibration and measurement in a single scan . Med. Phys. 39 ( 10 ), 6339 – 6350 ( 2012 ). Google Scholar CrossRef Search ADS PubMed 11 Poppinga , D. , Schoenfeld , A. , Doerner , K. , Blanck , O. , Harder , D. and Poppe , B. A new correction method serving to eliminate the parabola effect of flatbed scanners used in radiochromic film dosimetry . Med. Phys. 41 ( 2 ), 021707 ( 2014 ). Google Scholar CrossRef Search ADS PubMed 12 Reinhardt , S. , Hillbrand , M. , Wilkens , J. and Assmann , W. Comparison of Gafchromic EBT2 and EBT3 films for clinical photon and proton beams . Med. Phys. 39 ( 8 ), 5257 – 5262 ( 2012 ). Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: firstname.lastname@example.org 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.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera