- Email: [email protected]
Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy
Physica Medica xxx (2014) 1e6
Contents lists available at ScienceDirect
Physica Medica journal homepage: http://www.physicamedica.com
Technical Notes
Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy J. Camilleri a, b, *, J. Mazurier b, D. Franck b, P. Dudouet c, I. Latorzeff b, X. Franceries a, d, e a
INSERM, Imagerie cérébrale et handicaps neurologiques, UMR 825, F-31059 Toulouse, France Service de radiothérapie, groupe Oncorad-Garonne, Clinique Pasteur, L’Atrium, 1, rue de la petite-vitesse, 31000 Toulouse, France c Service d’oncologie, groupe Oncorad-Garonne, Clinique du Pont-de-Chaume, 330, avenue Marcel-Unal, 82000 Montauban, France d Université de Toulouse, UPS, INPT, LAPLACE Laboratoire Plasma et Conversion d’Energie, 118 route de Narbonne, F-31062 Toulouse Cedex 9, France e Université de Toulouse, UPS, Imagerie cérébrale et handicaps neurologiques, UMR 825, F-31059 Toulouse, France b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 October 2013 Received in revised form 14 February 2014 Accepted 19 February 2014 Available online xxx
The purpose of our work was to investigate the feasibility of using an EPID-based in-vivo dosimetry method initially designed for conformal fields on pelvic dynamic IMRT fields. The method enables a point dose delivered to the patient to be calculated from the transit signal acquired with an electronic portal imaging device (EPID). After defining a set of correction factors allowing EPID pixel values to be converted into absolute doses, several tests on homogeneous water-equivalent phantoms were performed to estimate the validity of the method in reference conditions. The effects of different treatment parameters, such as delivered dose, field size dependence and patient thickness were also studied. The model was first evaluated on a group of 53 patients treated by 3D conformal radiotherapy (3DCRT) and then on 92 patients treated by IMRT, both for pelvic cancers. For each measurement, the dose was reconstructed at the isocenter (DREC) and compared with the dose calculated by our treatment planning system (DTPS). Excellent agreement was found between DREC and DTPS for both techniques. For 3DCRT treatments, the mean deviation between DREC and DTPS for the 211 in-vivo dose verifications was equal to 1.0 2.2% (1SD). Concerning IMRT treatments, the averaged deviation for the 418 fields verified was equal to 0.3 2.6% (1SD) proving that the method is able to reconstruct a dose for dynamic IMRT pelvic fields. Based on these results, tolerance criteria and action levels were established before its implementation in clinical routine. Ó 2014 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.
Keywords: IMRT quality assurance In-vivo transit dosimetry Electronic portal imaging device
Introduction Intensity-modulated radiation therapy (IMRT) has significantly improved dose conformation to the target volume. Its ability to minimize the dose received by organs at risk close to the planned target volume has made this treatment modality standard in external radiotherapy. In 2011, at the Clinique Pasteur radiotherapy department, IMRT techniques were used for almost 25% of cases, more than half of which concerned pelvic locations. Due to the great complexity of the method and the high degree of accuracy required, the implementation of efficient and safe quality assurance (QA) is indispensable.
* Corresponding author. Service de Radiothérapie e Bât. Atrium, Clinique Pasteur, BP 27617, 31 076 Toulouse Cedex 3, France. Tel.: þ33(0)671679004. E-mail address: [email protected] (J. Camilleri).
Patient-specific pre-treatment QA is usually employed to check whether the delivered and planned doses coincide. Detector arrays, including electronic portal imaging devices (EPIDs), are suitable tools for such measurements. Although they were not originally intended for dosimetry, their characteristics, such as linearity of detector response, reproducibility and field size dependence, which are well reported in the literature [1e3], have led to the development of various algorithms enabling EPID images to be used to verify the delivery of IMRT fields. For this purpose, the calibrated EPID response can be either predicted or converted to dose for comparison with a treatment planning system (TPS) calculation [4e6]. Such methods verify whether the planned treatment is correctly delivered by the accelerator and can also highlight potential failures of the TPS calculation or the treatment delivery system. On the other hand, in-vivo dosimetry (i.e. during the treatment session) is a complex procedure for IMRT fields. Nevertheless, it remains the ultimate step to verify that the treatment session has
http://dx.doi.org/10.1016/j.ejmp.2014.02.003 1120-1797/Ó 2014 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
2
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
been successfully completed. An EPID-based in-vivo dosimetry method appears to be an interesting and efficient option for carrying out such measurements [7]. Two general approaches can be considered. The first consists of acquiring transit portal images during the treatment session and comparing them with a portal dose image computed at the EPID level [8,9]. The second one uses the transit portal images to reconstruct the dose within the patient [10,11]. In this way, 3D dose reconstruction becomes possible for both 3D conformal radiotherapy treatment (3DCRT) and IMRT [12,13]. The purpose of this study was to investigate the possibility to perform in-vivo dosimetry on dynamic IMRT pelvic plans with an EPID-based technique previously designed for conformal fields [14]. Several tests on a water-equivalent slab phantom were carried out to evaluate the method and also to study the influence of field size, phantom thickness and monitor units (MU) delivered on the dose reconstruction. Clinical results are then reported for two groups of patients, both treated for pelvic cancers. The first group (53 patients) was treated by 3DCRT and the second (92 patients) by dynamic IMRT. Material and methods Equipment Results presented here concern patients treated on a CLINAC 23 iX (Varian Medical System, Palo Alto, CA, USA) with a 120 multileaf collimator (MLC). The model was created for an aS-500Ô EPID and two photon beam energies which are 6 MV (QI ¼ 0.667) and 25 MV (QI ¼ 0.791). The detector was managed by the Image Acquisition System 3 (IAS3) software package. More details on the aS-500Ô technology can be found in references 1,2 and 3. EPID images were acquired using the integrated mode (corrected by dark field and flood field images) at a fixed source detector distance (SDD ¼ 150 cm) and for a dose rate equal to 300 UM$min1. Ionization measurements were made with a calibrated cylindrical ionization chamber, SEMIFLEX 31010, in combination with an electrometer, UNIDOS Universal Dosemeter (PTW-Freiburg, Freiburg, Germany), and converted into absolute doses according to the International Atomic Energy Agency (IAEA) dosimetry protocol TRS 398 [15]. Dose calculations were performed with TPS EclipseÔ version 10 using the Analytical Anisotropic Algorithm version 8.9.08 (Varian Medical System, Palo Alto, USA).
Eq. (1) and Fig. 1 summarize all the steps and parameters used in this study that are necessary to reconstruct the dose at dcalc in patient from SEPID. More details about the algorithm and its modeling can be found in Ref. [14].
DREC ðs; dcalc Þ ¼ SEPID ðs; t Þ CFEPID/IC ðs; t Þ TMR1 Transit ðs; t Þ FISQL TMRðs; dcalc Þ (1) Square field sizes of 4 4, 8 8, 10 10, 20 20 cm2 and absorber thicknesses of 5, 10, 15, 20, 30 and 40 cm were selected to create these data. These values were chosen so as to encompass a wide range of clinical situations. All measurements were taken by delivering 50 MU. To obtain an integrated signal, SEPID was obtained by multiplying the averaged absolute value of a 5 5 central pixel region of interest (PV5*5 CAX ) by the number of frames (AF) (Eq. (2)).
SEPID ¼ PV5*5 CAX AF
(2)
A model must be created for each EPID-Energy combination. The field size, the patient/phantom thickness and the depth where DREC is calculated have to be accurately determined for each field verified. The equivalent square field size for conformal fields was calculated by the TPS. For modulated fields, this parameter was determined by calculating the square root of the field aperture contour length delimited by the MLC. It has to be noted that this calculation field size method is only applicable for prostate IMRT fields which demonstrate relatively little modulation. Indeed, it gives an approximation to the effective equivalent square field size and thus, cannot be applied for the entire range of IMRT fields (like head and neck cancers) where increased modulation would cause the degree of approximation to be excessive. Patient thickness and depth were calculated manually on the patient CT. To take heterogeneities crossed by the beam during irradiation into consideration, the water-equivalent distances were always used. Then, a bilinear interpolation was used to calculate CFEPID/IC (s, t), TMRTransit (s, t) and TMR (s, dcalc). For all dose reconstructions, the deviation (x) between DREC and the dose calculated by the TPS (DTPS) was defined as percent difference with respect to DTPS. A distance-to-agreement (DTA) level was also defined as, for several reasons (high dose gradient region on beam central axis or possible detector sagging with gantry rotation), a shift of PV5*5 CAX from the central axis was possible. To account for this phenomenon, an overall tolerance of 2 mm was defined as DTA.
From EPID pixel value to patient dose The method used in this study was based on a back-projection method initially developed by Chang et al. and improved by François et al. for conformal fields [14,16]. It enables the on-axis patient dose to be calculated from the EPID signal (SEPID), at a chosen calculation depth (dcalc). To relate SEPID to an absolute dose in the patient, the method needs a set of correction factors created from EPID and ionization measurements for several square field sizes (s) and several absorber thicknesses (t) performed at the chosen SDD which are: i) A set of calibration factors (CFEPID/IC) to convert SEPID into an absolute transit dose in water at maximum depth (dmax). ii) A transmission function called the Transit Tissue Maximum Ratio (TMRTransit) to obtain the dose in water at dmax without patient. iii) A factor accounting for the inverse square law (FISQL) in order to obtain the dose in patient at dmax. iv) The Tissue Maximum Ratio (TMR) to finally calculate the dose in patient at dcalc.
Testing and validating the formalism on homogeneous phantom First, the formalism was tested by reconstructing the dose at the isocenter from the EPID signal on a water-equivalent slab phantom. Reference conditions (t ¼ 10 cm, s ¼ 10 10 cm2, dcalc ¼ 5 cm, delivered dose: 50 MU) were established for both the 6 and 25-MV photon beam energies. Ten successive EPID irradiations were carried out, respecting the same time between successive acquisitions (i.e. 30 s). This duration was chosen to recreate the effective time between two successive irradiations in clinical use and was not implemented due to ghosting concerns. Then, the influence of several treatment parameters was studied by irradiating the EPID under reference conditions while varying one parameter at a time. Three parameters were studied: i) The number of MU delivered (ranging from 5 to 200 MU). ii) The absorber thickness (ranging from 10 to 40 cm). iii) The field size (ranging from 5 5 to 20 20 cm2).
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
3
Figure 1. Schematic overview of the in-vivo dose reconstruction method used in this work. The EPID signal (SEPID) is first converted in an absolute transit dose in water (DTransit) at maximum depth (dmax) via a conversion factor (CFEPID/IC) defined for a field size (s) and a patient thickness (t). Then, the dose in water at dmax, without patient (DOpen) is obtained by using the inverse of the Transit Tissue Maximum Ratio (TMRTransit). Finally, the dose in patient (DREC) at a calculation depth (dcalc) is obtained by using an inverse square law factor (FISQL) and the Tissue Maximum Ratio (TMR) calculated for a (s, dcalc) pair.
A last test to assess the method’s ability to reconstruct the dose from modulated beams consisted of delivering five dynamic IMRT prostate plans (25 fields) to a 25-cm-thick water-equivalent slab phantom. For each plan, the prescribed dose per fraction was 2 Gy distributed over five beams. For these measurements, the gantry angle was set to 0 and DREC was calculated at the isocenter (dcalc ¼ 12.5 cm). Patient dose verification Over five months, the method was tested on two cohorts of patients treated by external radiotherapy, both for pelvic cancers. The first test concerned 53 patients treated by 3DCRT (3DCRT group) and was intended to study only the effects of patient presence on dose reconstruction. This first evaluation represented 211 in-vivo dose reconstructions. The second test concerned 92 patients treated by IMRT (IMRT group) and represented 418 fields. The two groups studied were treated on the same linear accelerator. For all in-vivo dose measurements, DREC was calculated at the isocenter and during the second treatment session. For the IMRT group, the treatment was always performed after daily CBCT matching. Results Dose reconstruction on homogeneous phantom Results concerning dose measurements under reference conditions for the ten successive EPID irradiations are reported in Table 1. The ten calculated dose deviations were averaged and showed good agreement between DREC and DTPS, since results were within 0.5% (0.2 0.3% for the 6-MV and 0.4 0.4% for the 25-MV photon beam energy). Moreover, for the chosen delay between measurements Table 1 Results concerning the 10 fields delivered under reference conditions for 6 and 25 MV photon beam energies.
Mean (%) Median (%) Min (%) Max (%) 1SD (%)
6 MV (n ¼ 10)
25 MV (n ¼ 10)
0.2 0.3 0.4 0.4 0.3
0.4 0.6 0.8 0.8 0.4
(which corresponded approximately to the time between two successive beams in a clinical situation), EPID response was quite reproducible. Results concerning the influence of the delivered dose, the field size and the phantom thickness on the EPID dose reconstruction are reported in Table 2. As can be seen in Fig. 2, both the linearity with dose above 20 MU and the known decrease in linearity at low MUs [1e3] were observed, resulting in an underestimation of the reconstructed dose (as much as 4.7% for the 6-MV photon beam energy). However, for a delivered dose ranging from 20 to 200 MU, the mean deviation between the reconstructed and the calculated dose was found to be equal to 0.0 0.6% (1SD) for the 6-MV and 0.1 0.4% (1SD) for the 25-MV photon beam. The influence of field size and phantom thickness is also represented in Fig. 2 and shows no major effects for field size values ranging from 5 5 to 20 20 cm2 and for patient thicknesses ranging from 10 to 40 cm. The histogram of ratios between DREC and DTPS for the five pelvic plans delivered to the homogeneous phantom is given in Fig. 3. For the 25 dose reconstructions, a mean deviation of 0.1 1.0% was found, which proved that the algorithm is able to reconstruct an absolute dose from an EPID transit signal for pelvic dynamic IMRT fields to similar accuracy as 3DCRT fields. Of the 117 doses reconstructed in the homogeneous phantom, more than 95% of percentage differences were within 2.5%, which demonstrates the accuracy of the method. Dose reconstruction on patients In-vivo dose reconstructions were performed on 145 patients, all treated for pelvic cancers. A mean deviation between DREC and DTPS Table 2 Results concerning the influence of the delivered dose, the field size and the phantom thickness on the EPID dose reconstruction.
Mean (%) Median (%) Min (%) Max (%) 1SD (%)
Monitor units (MU)
Absorber thickness (t)
Field size (s)
6 MV (n ¼ 22)
25 MV (n ¼ 22)
6 MV (n ¼ 7)
25 MV (n ¼ 7)
6 MV (n ¼ 7)
25 MV (n ¼ 7)
0.5 0.2 4.7 0.6 1.5
0.3 0.1 3.2 0.6 1.1
0.1 0.2 1.7 1.4 1.2
0.9 0.8 2.0 0.1 0.8
0.5 0.1 0.5 1.0 0.6
0.1 0.0 0.6 0.7 0.6
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
4
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
Figure 2. Evaluation of the dose reconstruction method on homogeneous water-equivalent phantom according the number of monitor units delivered during the irradiation, the absorber thickness (t) placed between the source and the EPID and the field size (s) for 6-MV (diamond) and 25-MV (triangle) photon beam energies. The deviation (x) between DREC and DTPS was defined as percent difference with respect to DTPS.
of 1.0 2.2% was found for the 53 patients treated by 3DCRT (211 verifications) and 0.3 2.6% for the 92 patients (418 verifications) treated by IMRT. Analysis of the histogram of ratios between DREC and DTPS calculated for the two cohorts showed a symmetrical
Figure 3. Repartition of ratios between doses reconstructed from EPID signal (DREC) and calculated by the TPS (DTPS) for 25 pelvic modulated fields reconstructed at the isocenter of an homogeneous water-equivalent slab phantom.
distribution with respect to the mean values, both for the 3DCRT group and the IMRT group (Fig. 4). Among the 629 dose verifications and after rejecting the false positive due to user errors like errors made by the radiation therapy technicians performing and analyzing measurements or during the determination of the individual patient parameters, 7 fields showed important deviations (3 for 3DCRT group and 4 for IMRT group). The major deviation (þ9.2%) occurred for a patient treated by 3DCRT and having a gas pocket in the rectum during the treatment session (viewable on the EPID images). Two other deviations were found on posterior beams (6.5% and 7.5%). In those cases, the presence of a gas pocket located in the rectum on the planned CT-scan was detected. This deviation between the patient’s anatomy and the anatomy according to the planned CT-scan generated an error in the determination of dcalc and t and thus in the dose reconstruction process. Notable dose differences (þ6.1% and þ6.2%) were observed for a patient who had lost weight and where his parameter t measured on the CBCT was reduce of almost 1 cm compared to the planned CT-scan. The last two high deviations occurred for 2 IMRT fields delivering low doses on beam central axis. In these cases, a large deviation was found in terms of percentage (6.1% for both cases) but represented minor differences in
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
5
Figure 4. Repartition of ratios between doses reconstructed from EPID signal (DREC) and calculated by the TPS (DTPS) for 53 patients (211 fields) treated by 3DCRT (left) and 92 patients (418 fields) treated by IMRT (right) both for pelvic cancers. Doses were reconstructed within patients at the isocenter.
terms of absolute dose since the difference between DREC and DTPS was always smaller than 1 cGy (i.e. 0.8 cGy and 0.6 cGy). Discussion Validation for pelvic IMRT treatments Several steps were necessary to validate the method. The first set of measurements performed on the water-equivalent slab phantom allowed the correction factors implemented in the algorithm to be validated for clinically relevant parameters (in terms of field sizes, dose delivered and patient/phantom thicknesses). Furthermore, the 117 EPID images analyzed showed that, even if a small ghosting effect was present, it did not significantly impact the dose reconstruction accuracy. Therefore, as other teams report [10,12], no correction for ghosting was made. The possibility to use the algorithm to reconstruct the dose from IMRT fields was then established by analyzing the 25 IMRT dose reconstructions in the homogeneous phantom. Francois et al. carried out a similar experiment for 46 3DCRT beams delivered in a polystyrene phantom [14]. The mean value of ratios between reconstructed and prescribed doses was equal to 0.998 0.026 which is very close to our IMRT results (Fig. 3) where a mean value equal to 1.001 0.010 was found, proving that the level of modulation concerning pelvic IMRT fields did not significantly affect the accuracy of the dose reconstruction process. Note that in our case, a lower standard deviation was found which can be explained by the difference of phantoms used. Indeed, in Ref. [14], the authors performed their measurements on a polystyrene phantom which is not, unlike our, water-equivalent. The observation of similarity in dose differences (in terms of magnitude and standard deviation) between the two groups of patients indicates that the dose reconstruction method responds similarly to both treatment techniques. Moreover, in most cases where notable dose differences were found, patient-related errors were the cause, which has also been observed with other EPIDbased dose reconstruction methods [17,18].
discrepancy. After mechanical problems or possible errors in the determination of the parameters necessary to calculate the dose have been eliminated, the patient’s CBCT is analyzed to find possible changes in patient position or anatomy compared with the planned CT-scan (e.g. patient weight loss, gas pocket in the rectum or possible motion of internal organ). Finally, to overcome deviations due to low doses delivered by the accelerator, a deviation on each individual field greater than 5% is accepted provided that it corresponds to a difference of less than 1 cGy between DREC and DTPS. This low absolute dose threshold was chosen in order to analyze all deviations higher than 5% to be sure to consider all possible errors that may occur in clinical routine. Now that our control of the process has improved, we plan to consider the total dose before accepting or rejecting a measurement. Conclusion In-vivo dose measurement is the last step of a quality assurance procedure to ensure that the dose delivered during treatment is in agreement with the prescribed one. In this work, a back-projection method based on transit dose EPID measurements initially developed for conformal beams was tested to verify IMRT pelvic plans. More than 700 fields have been checked in both 3DCRT and IMRT. Excellent agreement has been found with a maximum standard deviation of 2.2% for 3DCRT and 2.6% for IMRT proving that the algorithm is a suitable tool for pelvic IMRT in-vivo verifications. Currently, in-vivo dose measurement is used for all patients treated by IMRT for pelvic cancers in our department in order to detect potential errors that could affect the treatment quality. Acknowledgments The authors are indebted to Marianne Ducassou, Delphine Marre, Nicolas Mathy, Pauline Navarro and Daniel Zarate for their assistance with measurement and for fruitful discussions.
Tolerance criteria and action level References The study conducted over five months on patients also allowed us to established action levels and decision trees for integration into our clinical workflow. Based on these clinical results, a tolerance level of 5% (i.e. almost 2 SD of the 418 measurements on patients) for each individual beam between DREC and DTPS was established. If the deviation is greater than 5%, an investigation is conducted by the medical physicists to find the reason for this
[1] Greer PB, Popescu C. Dosimetric properties of an amorphous silicon electronic portal imaging device for verification of dynamic intensity modulated radiation therapy. Med Phys 2003;30:1618e27. [2] McCurdy BM, Luchka K, Pistorius S. Dosimetric investigation and portal dose image prediction using an amorphous silicon electronic portal imaging device. Med Phys 2001;28:911e24. [3] Winkler P, Hefner A, Georg D. Doseeresponse characteristics of an amorphous silicon EPID. Med Phys 2005;32:3095e105.
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
6
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
[4] van Esch A, Depuydt T, Huyskens DP. The use of an aSi-based EPID for routine absolute dosimetric pre-treatment verification of dynamic IMRT fields. Radiother Oncol 2004;71:223e34. [5] Richart J, Pujades MC, Perez-Calatayud J, Granero D, Ballester F, Rodriguez S, et al. QA of dynamic MLC based on portal dosimetry. Phys Med 2012;28:262e8. [6] Nelms B, Rasmussen K, Tome W. Evaluation of a fast method of EPID-based dosimetry for intensity-modulated radiation therapy. J Appl Clin Med Phys 2010;11:140e57. [7] van Elmpt W, McDermott LN, Nijsten SM, Wendling M, Lambin P, Mijnheer BJ. A literature review of electronic portal imaging for radiotherapy dosimetry. Radiother Oncol 2008;88:289e309. [8] Berry S, Sheu RD, Polvorosa C, Wuu CS. Implementation of EPID dosimetry based on a through-air dosimetry algorithm. Med Phys 2012;39:87e98. [9] Monville ME, Zuncic Z, Greer PB. Simulation of real-time EPID images during IMRT using Monte-Carlo. Phys Med; 2013. http://dx.doi.org/10.1016/j.ejmp. 2013.10.002. [10] Wendling M, Louwe R, McDermott LN, Sonke J-J, van Herk M, Mijnheer BJ. Accurate two-dimensional IMRT verification using a back-projection EPID dosimetry method. Med Phys 2006;33:259e73. [11] Fidanzio A, Cilla S, Greco F, Gargiulo L, Azario L, Sabatino D, et al. Generalized EPID calibration for in-vivo transit dosimetry. Phys Med 2011;27:30e8.
[12] Wendling M, McDermott LN, Mans A, Sonke J-J, van Herck M, Mijnheer BJ. A simple backprojection algorithm for 3D in vivo EPID dosimetry of IMRT treatments. Med Phys 2009;36:3310e21. [13] Mans A, Remeijer P, Olaciregui-Ruiz I, Wendling M, J-J Sonke, Mijnheer B, et al. 3D Dosimetric verification of volumetric-modulated arc therapy by portal dosimetry. Radiother Oncol 2010;94:181e7. [14] François P, Boissard P, Berger L, Mazal A. In vivo dose verification from back projection of a transit dose measurement on the central axis of photon beams. Phys Med 2011;27:1e10. [15] International Atomic Energy Agency. Absorbed dose determination in external beam radiotherapy: an international code of practice for dosimetry based on standards of absorbed dose to water. IAEA Technical Report Series 398. Vienna, Austria: IAEA; 2000. [16] Chang J, Mageras G. Relative profile and dose verification of intensitymodulated radiation therapy. Int J Radiat Oncol Biol Phys 2000;47:231e40. [17] Mans A, Wendling M, McDermott LN, Sonke J-J, Tielenburg R, Vijlbrief R, et al. Catching errors with in-vivo dosimetry. Med Phys 2010;37:2638e44. [18] Nijsten SM, Mijhneer BJ, Dekker AL, Lambin P, Minken A. Routine individualised patient dosimetry using electronic portal imaging devices. Radiother Oncol 2007;83:65e75.
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
Contents lists available at ScienceDirect
Physica Medica journal homepage: http://www.physicamedica.com
Technical Notes
Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy J. Camilleri a, b, *, J. Mazurier b, D. Franck b, P. Dudouet c, I. Latorzeff b, X. Franceries a, d, e a
INSERM, Imagerie cérébrale et handicaps neurologiques, UMR 825, F-31059 Toulouse, France Service de radiothérapie, groupe Oncorad-Garonne, Clinique Pasteur, L’Atrium, 1, rue de la petite-vitesse, 31000 Toulouse, France c Service d’oncologie, groupe Oncorad-Garonne, Clinique du Pont-de-Chaume, 330, avenue Marcel-Unal, 82000 Montauban, France d Université de Toulouse, UPS, INPT, LAPLACE Laboratoire Plasma et Conversion d’Energie, 118 route de Narbonne, F-31062 Toulouse Cedex 9, France e Université de Toulouse, UPS, Imagerie cérébrale et handicaps neurologiques, UMR 825, F-31059 Toulouse, France b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 October 2013 Received in revised form 14 February 2014 Accepted 19 February 2014 Available online xxx
The purpose of our work was to investigate the feasibility of using an EPID-based in-vivo dosimetry method initially designed for conformal fields on pelvic dynamic IMRT fields. The method enables a point dose delivered to the patient to be calculated from the transit signal acquired with an electronic portal imaging device (EPID). After defining a set of correction factors allowing EPID pixel values to be converted into absolute doses, several tests on homogeneous water-equivalent phantoms were performed to estimate the validity of the method in reference conditions. The effects of different treatment parameters, such as delivered dose, field size dependence and patient thickness were also studied. The model was first evaluated on a group of 53 patients treated by 3D conformal radiotherapy (3DCRT) and then on 92 patients treated by IMRT, both for pelvic cancers. For each measurement, the dose was reconstructed at the isocenter (DREC) and compared with the dose calculated by our treatment planning system (DTPS). Excellent agreement was found between DREC and DTPS for both techniques. For 3DCRT treatments, the mean deviation between DREC and DTPS for the 211 in-vivo dose verifications was equal to 1.0 2.2% (1SD). Concerning IMRT treatments, the averaged deviation for the 418 fields verified was equal to 0.3 2.6% (1SD) proving that the method is able to reconstruct a dose for dynamic IMRT pelvic fields. Based on these results, tolerance criteria and action levels were established before its implementation in clinical routine. Ó 2014 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.
Keywords: IMRT quality assurance In-vivo transit dosimetry Electronic portal imaging device
Introduction Intensity-modulated radiation therapy (IMRT) has significantly improved dose conformation to the target volume. Its ability to minimize the dose received by organs at risk close to the planned target volume has made this treatment modality standard in external radiotherapy. In 2011, at the Clinique Pasteur radiotherapy department, IMRT techniques were used for almost 25% of cases, more than half of which concerned pelvic locations. Due to the great complexity of the method and the high degree of accuracy required, the implementation of efficient and safe quality assurance (QA) is indispensable.
* Corresponding author. Service de Radiothérapie e Bât. Atrium, Clinique Pasteur, BP 27617, 31 076 Toulouse Cedex 3, France. Tel.: þ33(0)671679004. E-mail address: [email protected] (J. Camilleri).
Patient-specific pre-treatment QA is usually employed to check whether the delivered and planned doses coincide. Detector arrays, including electronic portal imaging devices (EPIDs), are suitable tools for such measurements. Although they were not originally intended for dosimetry, their characteristics, such as linearity of detector response, reproducibility and field size dependence, which are well reported in the literature [1e3], have led to the development of various algorithms enabling EPID images to be used to verify the delivery of IMRT fields. For this purpose, the calibrated EPID response can be either predicted or converted to dose for comparison with a treatment planning system (TPS) calculation [4e6]. Such methods verify whether the planned treatment is correctly delivered by the accelerator and can also highlight potential failures of the TPS calculation or the treatment delivery system. On the other hand, in-vivo dosimetry (i.e. during the treatment session) is a complex procedure for IMRT fields. Nevertheless, it remains the ultimate step to verify that the treatment session has
http://dx.doi.org/10.1016/j.ejmp.2014.02.003 1120-1797/Ó 2014 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
2
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
been successfully completed. An EPID-based in-vivo dosimetry method appears to be an interesting and efficient option for carrying out such measurements [7]. Two general approaches can be considered. The first consists of acquiring transit portal images during the treatment session and comparing them with a portal dose image computed at the EPID level [8,9]. The second one uses the transit portal images to reconstruct the dose within the patient [10,11]. In this way, 3D dose reconstruction becomes possible for both 3D conformal radiotherapy treatment (3DCRT) and IMRT [12,13]. The purpose of this study was to investigate the possibility to perform in-vivo dosimetry on dynamic IMRT pelvic plans with an EPID-based technique previously designed for conformal fields [14]. Several tests on a water-equivalent slab phantom were carried out to evaluate the method and also to study the influence of field size, phantom thickness and monitor units (MU) delivered on the dose reconstruction. Clinical results are then reported for two groups of patients, both treated for pelvic cancers. The first group (53 patients) was treated by 3DCRT and the second (92 patients) by dynamic IMRT. Material and methods Equipment Results presented here concern patients treated on a CLINAC 23 iX (Varian Medical System, Palo Alto, CA, USA) with a 120 multileaf collimator (MLC). The model was created for an aS-500Ô EPID and two photon beam energies which are 6 MV (QI ¼ 0.667) and 25 MV (QI ¼ 0.791). The detector was managed by the Image Acquisition System 3 (IAS3) software package. More details on the aS-500Ô technology can be found in references 1,2 and 3. EPID images were acquired using the integrated mode (corrected by dark field and flood field images) at a fixed source detector distance (SDD ¼ 150 cm) and for a dose rate equal to 300 UM$min1. Ionization measurements were made with a calibrated cylindrical ionization chamber, SEMIFLEX 31010, in combination with an electrometer, UNIDOS Universal Dosemeter (PTW-Freiburg, Freiburg, Germany), and converted into absolute doses according to the International Atomic Energy Agency (IAEA) dosimetry protocol TRS 398 [15]. Dose calculations were performed with TPS EclipseÔ version 10 using the Analytical Anisotropic Algorithm version 8.9.08 (Varian Medical System, Palo Alto, USA).
Eq. (1) and Fig. 1 summarize all the steps and parameters used in this study that are necessary to reconstruct the dose at dcalc in patient from SEPID. More details about the algorithm and its modeling can be found in Ref. [14].
DREC ðs; dcalc Þ ¼ SEPID ðs; t Þ CFEPID/IC ðs; t Þ TMR1 Transit ðs; t Þ FISQL TMRðs; dcalc Þ (1) Square field sizes of 4 4, 8 8, 10 10, 20 20 cm2 and absorber thicknesses of 5, 10, 15, 20, 30 and 40 cm were selected to create these data. These values were chosen so as to encompass a wide range of clinical situations. All measurements were taken by delivering 50 MU. To obtain an integrated signal, SEPID was obtained by multiplying the averaged absolute value of a 5 5 central pixel region of interest (PV5*5 CAX ) by the number of frames (AF) (Eq. (2)).
SEPID ¼ PV5*5 CAX AF
(2)
A model must be created for each EPID-Energy combination. The field size, the patient/phantom thickness and the depth where DREC is calculated have to be accurately determined for each field verified. The equivalent square field size for conformal fields was calculated by the TPS. For modulated fields, this parameter was determined by calculating the square root of the field aperture contour length delimited by the MLC. It has to be noted that this calculation field size method is only applicable for prostate IMRT fields which demonstrate relatively little modulation. Indeed, it gives an approximation to the effective equivalent square field size and thus, cannot be applied for the entire range of IMRT fields (like head and neck cancers) where increased modulation would cause the degree of approximation to be excessive. Patient thickness and depth were calculated manually on the patient CT. To take heterogeneities crossed by the beam during irradiation into consideration, the water-equivalent distances were always used. Then, a bilinear interpolation was used to calculate CFEPID/IC (s, t), TMRTransit (s, t) and TMR (s, dcalc). For all dose reconstructions, the deviation (x) between DREC and the dose calculated by the TPS (DTPS) was defined as percent difference with respect to DTPS. A distance-to-agreement (DTA) level was also defined as, for several reasons (high dose gradient region on beam central axis or possible detector sagging with gantry rotation), a shift of PV5*5 CAX from the central axis was possible. To account for this phenomenon, an overall tolerance of 2 mm was defined as DTA.
From EPID pixel value to patient dose The method used in this study was based on a back-projection method initially developed by Chang et al. and improved by François et al. for conformal fields [14,16]. It enables the on-axis patient dose to be calculated from the EPID signal (SEPID), at a chosen calculation depth (dcalc). To relate SEPID to an absolute dose in the patient, the method needs a set of correction factors created from EPID and ionization measurements for several square field sizes (s) and several absorber thicknesses (t) performed at the chosen SDD which are: i) A set of calibration factors (CFEPID/IC) to convert SEPID into an absolute transit dose in water at maximum depth (dmax). ii) A transmission function called the Transit Tissue Maximum Ratio (TMRTransit) to obtain the dose in water at dmax without patient. iii) A factor accounting for the inverse square law (FISQL) in order to obtain the dose in patient at dmax. iv) The Tissue Maximum Ratio (TMR) to finally calculate the dose in patient at dcalc.
Testing and validating the formalism on homogeneous phantom First, the formalism was tested by reconstructing the dose at the isocenter from the EPID signal on a water-equivalent slab phantom. Reference conditions (t ¼ 10 cm, s ¼ 10 10 cm2, dcalc ¼ 5 cm, delivered dose: 50 MU) were established for both the 6 and 25-MV photon beam energies. Ten successive EPID irradiations were carried out, respecting the same time between successive acquisitions (i.e. 30 s). This duration was chosen to recreate the effective time between two successive irradiations in clinical use and was not implemented due to ghosting concerns. Then, the influence of several treatment parameters was studied by irradiating the EPID under reference conditions while varying one parameter at a time. Three parameters were studied: i) The number of MU delivered (ranging from 5 to 200 MU). ii) The absorber thickness (ranging from 10 to 40 cm). iii) The field size (ranging from 5 5 to 20 20 cm2).
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
3
Figure 1. Schematic overview of the in-vivo dose reconstruction method used in this work. The EPID signal (SEPID) is first converted in an absolute transit dose in water (DTransit) at maximum depth (dmax) via a conversion factor (CFEPID/IC) defined for a field size (s) and a patient thickness (t). Then, the dose in water at dmax, without patient (DOpen) is obtained by using the inverse of the Transit Tissue Maximum Ratio (TMRTransit). Finally, the dose in patient (DREC) at a calculation depth (dcalc) is obtained by using an inverse square law factor (FISQL) and the Tissue Maximum Ratio (TMR) calculated for a (s, dcalc) pair.
A last test to assess the method’s ability to reconstruct the dose from modulated beams consisted of delivering five dynamic IMRT prostate plans (25 fields) to a 25-cm-thick water-equivalent slab phantom. For each plan, the prescribed dose per fraction was 2 Gy distributed over five beams. For these measurements, the gantry angle was set to 0 and DREC was calculated at the isocenter (dcalc ¼ 12.5 cm). Patient dose verification Over five months, the method was tested on two cohorts of patients treated by external radiotherapy, both for pelvic cancers. The first test concerned 53 patients treated by 3DCRT (3DCRT group) and was intended to study only the effects of patient presence on dose reconstruction. This first evaluation represented 211 in-vivo dose reconstructions. The second test concerned 92 patients treated by IMRT (IMRT group) and represented 418 fields. The two groups studied were treated on the same linear accelerator. For all in-vivo dose measurements, DREC was calculated at the isocenter and during the second treatment session. For the IMRT group, the treatment was always performed after daily CBCT matching. Results Dose reconstruction on homogeneous phantom Results concerning dose measurements under reference conditions for the ten successive EPID irradiations are reported in Table 1. The ten calculated dose deviations were averaged and showed good agreement between DREC and DTPS, since results were within 0.5% (0.2 0.3% for the 6-MV and 0.4 0.4% for the 25-MV photon beam energy). Moreover, for the chosen delay between measurements Table 1 Results concerning the 10 fields delivered under reference conditions for 6 and 25 MV photon beam energies.
Mean (%) Median (%) Min (%) Max (%) 1SD (%)
6 MV (n ¼ 10)
25 MV (n ¼ 10)
0.2 0.3 0.4 0.4 0.3
0.4 0.6 0.8 0.8 0.4
(which corresponded approximately to the time between two successive beams in a clinical situation), EPID response was quite reproducible. Results concerning the influence of the delivered dose, the field size and the phantom thickness on the EPID dose reconstruction are reported in Table 2. As can be seen in Fig. 2, both the linearity with dose above 20 MU and the known decrease in linearity at low MUs [1e3] were observed, resulting in an underestimation of the reconstructed dose (as much as 4.7% for the 6-MV photon beam energy). However, for a delivered dose ranging from 20 to 200 MU, the mean deviation between the reconstructed and the calculated dose was found to be equal to 0.0 0.6% (1SD) for the 6-MV and 0.1 0.4% (1SD) for the 25-MV photon beam. The influence of field size and phantom thickness is also represented in Fig. 2 and shows no major effects for field size values ranging from 5 5 to 20 20 cm2 and for patient thicknesses ranging from 10 to 40 cm. The histogram of ratios between DREC and DTPS for the five pelvic plans delivered to the homogeneous phantom is given in Fig. 3. For the 25 dose reconstructions, a mean deviation of 0.1 1.0% was found, which proved that the algorithm is able to reconstruct an absolute dose from an EPID transit signal for pelvic dynamic IMRT fields to similar accuracy as 3DCRT fields. Of the 117 doses reconstructed in the homogeneous phantom, more than 95% of percentage differences were within 2.5%, which demonstrates the accuracy of the method. Dose reconstruction on patients In-vivo dose reconstructions were performed on 145 patients, all treated for pelvic cancers. A mean deviation between DREC and DTPS Table 2 Results concerning the influence of the delivered dose, the field size and the phantom thickness on the EPID dose reconstruction.
Mean (%) Median (%) Min (%) Max (%) 1SD (%)
Monitor units (MU)
Absorber thickness (t)
Field size (s)
6 MV (n ¼ 22)
25 MV (n ¼ 22)
6 MV (n ¼ 7)
25 MV (n ¼ 7)
6 MV (n ¼ 7)
25 MV (n ¼ 7)
0.5 0.2 4.7 0.6 1.5
0.3 0.1 3.2 0.6 1.1
0.1 0.2 1.7 1.4 1.2
0.9 0.8 2.0 0.1 0.8
0.5 0.1 0.5 1.0 0.6
0.1 0.0 0.6 0.7 0.6
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
4
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
Figure 2. Evaluation of the dose reconstruction method on homogeneous water-equivalent phantom according the number of monitor units delivered during the irradiation, the absorber thickness (t) placed between the source and the EPID and the field size (s) for 6-MV (diamond) and 25-MV (triangle) photon beam energies. The deviation (x) between DREC and DTPS was defined as percent difference with respect to DTPS.
of 1.0 2.2% was found for the 53 patients treated by 3DCRT (211 verifications) and 0.3 2.6% for the 92 patients (418 verifications) treated by IMRT. Analysis of the histogram of ratios between DREC and DTPS calculated for the two cohorts showed a symmetrical
Figure 3. Repartition of ratios between doses reconstructed from EPID signal (DREC) and calculated by the TPS (DTPS) for 25 pelvic modulated fields reconstructed at the isocenter of an homogeneous water-equivalent slab phantom.
distribution with respect to the mean values, both for the 3DCRT group and the IMRT group (Fig. 4). Among the 629 dose verifications and after rejecting the false positive due to user errors like errors made by the radiation therapy technicians performing and analyzing measurements or during the determination of the individual patient parameters, 7 fields showed important deviations (3 for 3DCRT group and 4 for IMRT group). The major deviation (þ9.2%) occurred for a patient treated by 3DCRT and having a gas pocket in the rectum during the treatment session (viewable on the EPID images). Two other deviations were found on posterior beams (6.5% and 7.5%). In those cases, the presence of a gas pocket located in the rectum on the planned CT-scan was detected. This deviation between the patient’s anatomy and the anatomy according to the planned CT-scan generated an error in the determination of dcalc and t and thus in the dose reconstruction process. Notable dose differences (þ6.1% and þ6.2%) were observed for a patient who had lost weight and where his parameter t measured on the CBCT was reduce of almost 1 cm compared to the planned CT-scan. The last two high deviations occurred for 2 IMRT fields delivering low doses on beam central axis. In these cases, a large deviation was found in terms of percentage (6.1% for both cases) but represented minor differences in
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
5
Figure 4. Repartition of ratios between doses reconstructed from EPID signal (DREC) and calculated by the TPS (DTPS) for 53 patients (211 fields) treated by 3DCRT (left) and 92 patients (418 fields) treated by IMRT (right) both for pelvic cancers. Doses were reconstructed within patients at the isocenter.
terms of absolute dose since the difference between DREC and DTPS was always smaller than 1 cGy (i.e. 0.8 cGy and 0.6 cGy). Discussion Validation for pelvic IMRT treatments Several steps were necessary to validate the method. The first set of measurements performed on the water-equivalent slab phantom allowed the correction factors implemented in the algorithm to be validated for clinically relevant parameters (in terms of field sizes, dose delivered and patient/phantom thicknesses). Furthermore, the 117 EPID images analyzed showed that, even if a small ghosting effect was present, it did not significantly impact the dose reconstruction accuracy. Therefore, as other teams report [10,12], no correction for ghosting was made. The possibility to use the algorithm to reconstruct the dose from IMRT fields was then established by analyzing the 25 IMRT dose reconstructions in the homogeneous phantom. Francois et al. carried out a similar experiment for 46 3DCRT beams delivered in a polystyrene phantom [14]. The mean value of ratios between reconstructed and prescribed doses was equal to 0.998 0.026 which is very close to our IMRT results (Fig. 3) where a mean value equal to 1.001 0.010 was found, proving that the level of modulation concerning pelvic IMRT fields did not significantly affect the accuracy of the dose reconstruction process. Note that in our case, a lower standard deviation was found which can be explained by the difference of phantoms used. Indeed, in Ref. [14], the authors performed their measurements on a polystyrene phantom which is not, unlike our, water-equivalent. The observation of similarity in dose differences (in terms of magnitude and standard deviation) between the two groups of patients indicates that the dose reconstruction method responds similarly to both treatment techniques. Moreover, in most cases where notable dose differences were found, patient-related errors were the cause, which has also been observed with other EPIDbased dose reconstruction methods [17,18].
discrepancy. After mechanical problems or possible errors in the determination of the parameters necessary to calculate the dose have been eliminated, the patient’s CBCT is analyzed to find possible changes in patient position or anatomy compared with the planned CT-scan (e.g. patient weight loss, gas pocket in the rectum or possible motion of internal organ). Finally, to overcome deviations due to low doses delivered by the accelerator, a deviation on each individual field greater than 5% is accepted provided that it corresponds to a difference of less than 1 cGy between DREC and DTPS. This low absolute dose threshold was chosen in order to analyze all deviations higher than 5% to be sure to consider all possible errors that may occur in clinical routine. Now that our control of the process has improved, we plan to consider the total dose before accepting or rejecting a measurement. Conclusion In-vivo dose measurement is the last step of a quality assurance procedure to ensure that the dose delivered during treatment is in agreement with the prescribed one. In this work, a back-projection method based on transit dose EPID measurements initially developed for conformal beams was tested to verify IMRT pelvic plans. More than 700 fields have been checked in both 3DCRT and IMRT. Excellent agreement has been found with a maximum standard deviation of 2.2% for 3DCRT and 2.6% for IMRT proving that the algorithm is a suitable tool for pelvic IMRT in-vivo verifications. Currently, in-vivo dose measurement is used for all patients treated by IMRT for pelvic cancers in our department in order to detect potential errors that could affect the treatment quality. Acknowledgments The authors are indebted to Marianne Ducassou, Delphine Marre, Nicolas Mathy, Pauline Navarro and Daniel Zarate for their assistance with measurement and for fruitful discussions.
Tolerance criteria and action level References The study conducted over five months on patients also allowed us to established action levels and decision trees for integration into our clinical workflow. Based on these clinical results, a tolerance level of 5% (i.e. almost 2 SD of the 418 measurements on patients) for each individual beam between DREC and DTPS was established. If the deviation is greater than 5%, an investigation is conducted by the medical physicists to find the reason for this
[1] Greer PB, Popescu C. Dosimetric properties of an amorphous silicon electronic portal imaging device for verification of dynamic intensity modulated radiation therapy. Med Phys 2003;30:1618e27. [2] McCurdy BM, Luchka K, Pistorius S. Dosimetric investigation and portal dose image prediction using an amorphous silicon electronic portal imaging device. Med Phys 2001;28:911e24. [3] Winkler P, Hefner A, Georg D. Doseeresponse characteristics of an amorphous silicon EPID. Med Phys 2005;32:3095e105.
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003
6
J. Camilleri et al. / Physica Medica xxx (2014) 1e6
[4] van Esch A, Depuydt T, Huyskens DP. The use of an aSi-based EPID for routine absolute dosimetric pre-treatment verification of dynamic IMRT fields. Radiother Oncol 2004;71:223e34. [5] Richart J, Pujades MC, Perez-Calatayud J, Granero D, Ballester F, Rodriguez S, et al. QA of dynamic MLC based on portal dosimetry. Phys Med 2012;28:262e8. [6] Nelms B, Rasmussen K, Tome W. Evaluation of a fast method of EPID-based dosimetry for intensity-modulated radiation therapy. J Appl Clin Med Phys 2010;11:140e57. [7] van Elmpt W, McDermott LN, Nijsten SM, Wendling M, Lambin P, Mijnheer BJ. A literature review of electronic portal imaging for radiotherapy dosimetry. Radiother Oncol 2008;88:289e309. [8] Berry S, Sheu RD, Polvorosa C, Wuu CS. Implementation of EPID dosimetry based on a through-air dosimetry algorithm. Med Phys 2012;39:87e98. [9] Monville ME, Zuncic Z, Greer PB. Simulation of real-time EPID images during IMRT using Monte-Carlo. Phys Med; 2013. http://dx.doi.org/10.1016/j.ejmp. 2013.10.002. [10] Wendling M, Louwe R, McDermott LN, Sonke J-J, van Herk M, Mijnheer BJ. Accurate two-dimensional IMRT verification using a back-projection EPID dosimetry method. Med Phys 2006;33:259e73. [11] Fidanzio A, Cilla S, Greco F, Gargiulo L, Azario L, Sabatino D, et al. Generalized EPID calibration for in-vivo transit dosimetry. Phys Med 2011;27:30e8.
[12] Wendling M, McDermott LN, Mans A, Sonke J-J, van Herck M, Mijnheer BJ. A simple backprojection algorithm for 3D in vivo EPID dosimetry of IMRT treatments. Med Phys 2009;36:3310e21. [13] Mans A, Remeijer P, Olaciregui-Ruiz I, Wendling M, J-J Sonke, Mijnheer B, et al. 3D Dosimetric verification of volumetric-modulated arc therapy by portal dosimetry. Radiother Oncol 2010;94:181e7. [14] François P, Boissard P, Berger L, Mazal A. In vivo dose verification from back projection of a transit dose measurement on the central axis of photon beams. Phys Med 2011;27:1e10. [15] International Atomic Energy Agency. Absorbed dose determination in external beam radiotherapy: an international code of practice for dosimetry based on standards of absorbed dose to water. IAEA Technical Report Series 398. Vienna, Austria: IAEA; 2000. [16] Chang J, Mageras G. Relative profile and dose verification of intensitymodulated radiation therapy. Int J Radiat Oncol Biol Phys 2000;47:231e40. [17] Mans A, Wendling M, McDermott LN, Sonke J-J, Tielenburg R, Vijlbrief R, et al. Catching errors with in-vivo dosimetry. Med Phys 2010;37:2638e44. [18] Nijsten SM, Mijhneer BJ, Dekker AL, Lambin P, Minken A. Routine individualised patient dosimetry using electronic portal imaging devices. Radiother Oncol 2007;83:65e75.
Please cite this article in press as: Camilleri J, et al., Clinical results of an EPID-based in-vivo dosimetry method for pelvic cancers treated by intensity-modulated radiation therapy, Physica Medica (2014), http://dx.doi.org/10.1016/j.ejmp.2014.02.003