Medical Dosimetry 37 (2012) Medical Dosimetry. journal homepage:

Medical Dosimetry 37 (2012) 53-60 Medical Dosimetry journal homepage: www.meddos.org Characterization of responses of 2d array seven29 detector and its combined use with octavius phantom for the patient-specific quality assurance in rapidarc treatment delivery S. A. Syamkumar, M.Sc., Sriram Padmanabhan, M.Sc., Prabakar Sukumar, M.Sc., and Vivekanandan Nagarajan, Ph.D. Department of Medical Physics, Cancer Institute (WIA), Chennai, India ARTICLE INFO ABSTRACT Article history: Received 5 July 2010 Accepted 24 December 2010 Keywords: 2D Seven29 array RapidArc Quality assurance A commercial 2D array seven29 detector has been characterized and its performance has been evaluated. 2D array ionization chamber equipped with 729 ionization chambers uniformly arranged in a27 27 matrix with an active area of 27 27 cm 2 was used for the study. An octagon-shaped phantom (Octavius Phantom) with a central cavity is used to insert the 2D ion chamber array. All measurements were done with a linear accelerator. The detector dose linearity, reproducibility, output factors, dose rate, source to surface distance (SSD), and directional dependency has been studied. The performance of the 2D array, when measuring clinical dose maps, was also investigated. For pretreatment quality assurance, 10 different RapidArc plans conforming to the clinical standards were selected. The 2D array demonstrates an excellent short-term output reproducibility. The long-term reproducibility was found to be within 1% over a period of 5 months. Output factor measurements for the central chamber of the array showed no considerable deviation from ion chamber measurements. We found that the 2D array exhibits directional dependency for static fields. Measurement of beam profiles and wedge-modulated fields with the 2D array matched very well with the ion chamber measurements in the water phantom. The study shows that 2D array seven29 is a reliable and accurate dosimeter and a useful tool for quality assurance. The combination of the 2D array with the Octavius phantom proved to be a fast and reliable method for pretreatment verification of rotational treatments. 2012 American Association of Medical Dosimetrists. Introduction Reprint requests to: S. A. Syamkumar, M.Sc., Department of Medical Physics, Cancer Institute (WIA), Sardar Patel Road, Adyar, Chennai, Tamil Nadu 600036, India. E-mail: skppm@rediffmail.com The verification of radiotherapy treatment plans is a very important step in complex radiotherapy techniques because the primary goal of radiation therapy is to deliver doses of ionizing radiation to a target volume while minimizing the dose to critical organs and healthy tissues. Ionization chamber array has become the standard device for quality assurance measurements in modern radiotherapy. Although radiographic film proved overall to be the most practical and cost-effective method, there are some difficulties with using the verification film. The film is affected by the processor characteristics at development time, creating the need for producing a film calibration curve for each quality assurance to be performed, even for films from the same batch. The verification film response is energy dependent and can cause dosimetry errors in measuring changing energy spectrum fields. Also film-based measurements consume more time compared with 2D array measurements. The 2D array ionization chamber devices are easy to use and provide quality assurance results while measurements are being performed. The complexity of treatment delivery has been increased recently by the implementation of intensity-modulated radiation therapy (IMRT), intensity-modulated arc therapy, and tomotherapy. 1 5 It is now possible to produce small, irregular fields and dose shaping by the use of multileaf collimators (MLCs). Studies by Zelefsky et al., Kam et al., and Krueger et al. have shown clinical advantages in these new planning and delivery techniques. Various studies have been performed with the 2D array seven29 for measuring IMRT delivery verification. 6 10 This complexity of clinical treatment planning and delivery raises the need for more accurate dose measurement and verification systems. In this study, we aimed to characterize the 2D array seven29 detector for its response, such as linearity, reproducibility, output fac- 0958-3947/$ see front matter Copyright 2012 American Association of Medical Dosimetrists doi:10.1016/j.meddos.2010.12.013

54 S. A. Syamkumar et al. / Medical Dosimetry 37 (2012) 53-60 Fig. 1. Standard measurement setup for the 2D ion chamber array. The array is sandwiched between a build-up and backscatter material. tors dependency, dose-rate dependency, and sensitivity for photon beams. In particular, we compared the 2D array responses for static as well as rotational deliveries. Also, the performance of the 2D array ion chamber for measuring clinical dose maps has been studied. Pretreatment patient-specific quality assurance for 10 different RapidArc (Varian Medical Systems, Palo Alto, CA) cases were done using a 2D array combined with Octavius phantom and analyzed using the PTW Verisoft software (Freiburg, Germany). Materials and Methods All measurements were done on Varian Clinac 2100 C/D (Varian Medical Systems, Palo, Alto, CA) with 6- and 15-MV photons. The radiation detector, 2D seven29 ion chamber array (T10024) model (PTW) was used in this work. The 2D array consists of 729 air-vented cubic ionization chambers uniformly arranged in a 27 27 matrix with an active area of 27 cm 2. The ionization chambers in the 2D array are separated by 0.5 cm and the centerto-center distance between 2 adjacent chambers is 1 cm. Each chamber has a volume of 0.125 cm 3. The linear dimensions of the 2D array are 2.2 30.0 42.0 cm 3. The detector is a pixel-segmented ionization chamber whose main features are 2D read-out capability, large detection area, good homogeneity, and dead time-free read-out. The 2D arrays are operated at a chamber voltage of 400 V. The reference point of the detector is located at 0.5 cm behind the 2D array surface. The wall material is made up of graphite and the material surrounding the vented ionization chamber is polymethyl methacrylate. Dose and dose rate mode is available for the measurements. The measurement ranges for absolute dose is 200 mgy to 1000 Gy, and for dose rate measurements from 500 mgy/min to 10 Gy/min as specified by the manufacturer. The 2D array is calibrated for absolute dosimetry in a Co60 photon beam at the PTW secondary standard dosimetry laboratory. Throughout this work, the detector array was used in absolute dose measuring mode. The dose rate of the machine is kept at 300 MU/min, which is the pulse rate mode mostly used in clinical practice, except for verification of the dose rate dependency study. PTW matrix scan software was used to acquire the data from the 2D array detector. Before all measurements, the 2D array was calibrated by delivering a known dose of radiation for a10 10-cm 2 field size under reference conditions. To account for the buildup and backscatter, the 2D array ion chamber array is sandwiched between the virtual water phantom (Medtec Inc., Orange City, IA). The effective point of measurement of the 2D array was kept at 5 cm depth from the surface of the virtual water phantom and the source-to-surface distance (SSD) at 95 cm (Fig. 1). The detector is placed so the central axis of beam passes through the central ion chamber (14 14). All results were compared with independent measurements using ionization chambers. Characterization of 2D array seven29 Verification of linearity The dose linearity was evaluated by irradiating the 2D array for a field size of 10 10 cm 2 for 6-MV and 15-MV photon beams. The output of 2D array for various monitor units (MUs) was determined. Verification of reproducibility The reproducibility is the percentage difference between consecutive measurements for the same radiation dose. The performance of 2D array seven29 was measured Fig. 2. Verification plan window in Eclipse treatment planning system (version 8.6) for the pretreatment patient quality assurance with the 2D array and Octavius phantom.

S. A. Syamkumar et al. / Medical Dosimetry 37 (2012) 53-60 55 Fig. 5. Output factor comparison between 2D array and pinpoint chamber for 6-MV and 15-MV photons. Fig. 3. Setup for the pretreatment quality assurance tests for RapidArc treatment delivery. 2D array inserted inside the Octavius phantom. SSD has been kept at 84 cm. to verify short-term, over a period of hours, and long-term reproducibility, over a period of 5 months. The output was measured by delivering 100 MU for a fixed field size of 15 15 cm 2. The measurement was repeated for 10 readings. Verification of output factor dependency The response of a small volume ionization chamber for small field size is of greater importance because of its potential applications for the verification of IMRT plans. 11 The performance of 2D array for radiation output was measured by delivering 100 MU for various field sizes ranging from 2 2cm 2 to 27 27 cm 2 for 6-MV and 15-MV photon beams, respectively. Verification of dose rate dependency The response of 2D array toward dose rate was measured and compared for 6-MV and 15-MV X-rays. The detector was irradiated by delivering 200 cgy for a 10 10-cm 2 field size at various dose rates (100, 200, 300, 400, 500, and 600 MU/min). Verification of SSD dependency The SSD dependency was studied for 6-MV and 15-MV photon beams. For 100 MU and 10 10-cm 2 field size, the doses were measured for different source-todetector distance. Verification of directional dependency The directional dependence of the 2D array detector was measured as a function of beam angles. An octagon-shaped phantom (Octavius phantom) having a width of 32 cm and a length of 32 cm made of polystyrene (physical density 1.04 g/cm 3, relative electron density 1.00) with a central cavity (30 30 2.2 cm 3 ) was used to insert the 2D ion chamber array for the verification of directional dependency. The source to detector effective point of measurement was kept at 100 cm and the SSD at 84 cm. The directional dependency was studied for static delivery by keeping the 2D array vertically inside the Octavius phantom to avoid the radiation beam passing through the couch. For the field size of 15 15 cm 2, the readings were acquired by delivering 100 MU for gantry angles from 90 270, with a gantry angle difference of 15. Therefore, in this study gantry 90 corresponds to the orthogonal beam incidence from the front side of the array and gantry 270 corresponds to the irradiation through the rear of the array. Also open dynamic arcs (gantry angle from 240 120 in clockwise direction) for 10 10-cm 2 and 20 20-cm 2 field sizes were delivered to check the directional dependency for rotational delivery. Clinical applications The open (20 20 cm 2 ) and wedge field profiles were measured with a 2D array. Also, complex MLC test patterns, such as chair pattern, sweeping field, and split fields, were done. The 2D array measured fluence was compared with the treatment planning system (TPS) calculated using the gamma analysis method. Pretreatment quality assurance of RapidArc using 2D array and Octavius phantom Volumetric arc modulation using RapidArc (Varian Medical Systems) is a method for delivering IMRT precisely using rotational beams within a very short period than the conventional IMRT. 12,13 RapidArc rotates 360 around the patient, enabling very small beams with varying intensity to be aimed at the tumor from multiple angles. Unlike helical IMRT treatments or other forms of radiation therapy, with RapidArc the radiation treatment delivered to the patient can be modulated continuously throughout treatment because the beam is on even when the gantry is moving. 4,5 In the RapidArc approach, both the treatment planning and linac systems incorporate the following capabilities: variable dose rate, variable gantry speed, Dynamic Multi Leaf Collimator (DMLC) movement. Because the RapidArc delivery involves complex treatment delivery procedures like variable dose rate and variable gantry speed, the patient-specific quality assurance for each patient should be done. RapidArc treatment planning The ability of the detector array for measuring planar dose distributions has been evaluated for different RapidArc cases. Ten different RapidArc plans (head and neck, esophagus, cervix) conforming to the clinical standards were selected for this study. All planning was done using the Eclipse planning system (version 8.6, Varian Medica Systems) using the AAA algorithm 14,15 (analytical anisotropic algorithm). The optimization is based on the so-called progressive resolution optimization (PRO) algorithm. Verification plans were done for all the cases with the 2D array and the Octavius phantom (Fig. 2). The 2D isocenter dose planes calculated in the TPS were exported to the Verisoft software for evaluation using the Gamma analysis method proposed by Low et al. 16 The acceptance criteria of 3 mm for the distance to agreement (DTA) and dose difference tolerance level of 3% were chosen. Also the percentage of the evaluated dose points passing the gamma index was kept at a limit 95%. The experimental setup is shown in Fig. 3. The values acquired using matrix scan software correspond to the dose in Gy for each ionization chamber. Verisoft software averages the calculated dose distribution automatically if a 2D array data file is loaded. Results Verification of linearity Fig. 4. Linearity test for 6-MV and 15-MV photon beams. Values correspond to the central axis ionization chamber dose. From the linear response curve (Fig. 4), the results show that the detector has a high degree of linearity within the range of 2 500 MU with a linearity coefficient R of 1.0 for both energies. Also the least squares fit shows that the linear relationship between MU and ionization chamber response is good for the lowest delivered doses.

56 S. A. Syamkumar et al. / Medical Dosimetry 37 (2012) 53-60 Fig. 6. Dose rate response curve for 6-MV and 15-MV photon energy. Verification of reproducibility The reproducibility of the measurements within each set is excellent. Also the variation of response from chamber to chamber is 1%. The 2D array demonstrates excellent short-term output reproducibility with a maximum standard deviation of 0.1%. The long-term reproducibility was found to be within 1% (standard deviation) over a period of 5 months for both 6-MV and 15-MV energies. This agrees with the results published by Spezi et al. 17 All measurements provide a record of the relative sensitivity of each ionization chamber with respect to the central ionization chamber in the array, presuming that the beam flatness and symmetry are unchanged during this period. Verification of output factor dependency Figure 5 shows the field size dependent output factor curve of the central ionization chamber of the 2D array for 6-MV and 15-MV photon energy. The point doses were verified with an ionization chamber in a water phantom for the same measurement setup. Output factor measurements for the central chamber of the array showed no considerable deviation from ion chamber measurements for bigger field sizes. The small deviation observed in the higher field size is because of the difference in the phantom used for measurements. Maximum variation for chamber and 2D array measurements is found to be 1.3% for 6 MV and 1.2% for 15 MV. For field sizes below 3 3cm 2, pinpoint chambers show the correct values, whereas the 2D array ionization chambers, because of the volume effects, 18,19 tend to slightly underestimate the true output factor. Thus it can be concluded that the sensitivity of the ionization chambers in the 2D array does not have any increased energy dependence that could increase their response to scattered photons. Also the scatter property of the 2D array is nearly equal to that of water. Verification of dose rate dependency Figure 6 shows the dose rate dependency curve from 100 600 MU/min for 6-MV and 15-MV photon energies. The results show that the 2D array have high dose rate independent response for the dose rates ranging from 100 600 MU/min with a standard deviation of 0.7% for 6-MV and 0.5% for 15-MV photon energies between the measured values. The results were compared with a 0.6-cc ionization chamber with a maximum variation of 0.5% with 6 MV and 0.9% with 15 MV. Verification of SSD dependency The responses of the detectors as a function of SSD for 6-MV and 15-MV photon beams are displayed in Fig. 7. The results were com- Fig. 7. SSD response curve for 6-MV and 15-MV. 2D array values have been compared with ion chamber measurements.

S. A. Syamkumar et al. / Medical Dosimetry 37 (2012) 53-60 57 Fig. 8. Directional dependency for 6-MV and 15-MV static photon beams. pared with those obtained using a farmer chamber for the same measurement setup. The 2D array agrees with the ionization chamber measurement within 1% for the range of SSDs performed in this study. Verification of directional dependency The detector array shows some angular dependency when it is irradiated from the lateral as well as the bottom side. The percentage variation for TPS calculated and 2D array measured static fields was found to be 0.7% when the array is irradiated from the front side for 6-MV photon energy and 1.49% deviation for 15-MV photon energy. Also 5.12% and 6.24% was observed for 6 MV and 15 MV, respectively, whereas the 2D array was irradiated parallel to the beam axis. The 2D array predicts slightly less dose when it is irradiated through the bottom side. The deviation was found to be 4.9% for 6-MVand 5.41% for 15-MV photon energies. Figure 8 shows the directional dependency of 2D array vs. semiflex ionization chamber measured for static delivery. The deviation is found to be 0.7% and 1.17% for 6 MV and 15 MV, respectively, when the array is irradiated from the front side. A maximum variation of 4.65% for 6 MV and 5.14% for 15 MV were observed when the 2D array was irradiated parallel to the beam axis. An absolute dose deviation of 1.02% for 6 MV and 1.87% for 15 MV was observed while the beam incident passing through the posterior side of the array. Results are comparable with the published data by Van Esch et al. 13 The results for directional dependency for 6-MV and 15-MV arc deliveries are shown in Fig. 9. The measured open arcs were analyzed with the TPS planned using the Gamma analysis method. For all the arc deliveries, the TPS and the measured value agree well within 3-mm DTA, 3% dose difference tolerance level (DD), for 95% of the evaluated dose points using the gamma analysis method. The points where the gamma analysis fails was mainly observed in the penumbral region. Clinical applications Figure 10 shows the profiles of open and different wedge-modulated fields for 6-MV photon beams. The results have been compared with a semiflex ionization chamber. All measured data were normalized to the central axis. The open beam profile comparison shows very good agreement with ionization chamber measurements. Similarly, wedge field profile results match very well with the ion chamber measurements for all 4 wedges (15, 30, 45, 60 ). Also chair pattern and split-fields test for MLC positions and sweep-field tests for MLC performance shows good agreement with the calculated fluence using TPS. Figure 11, a and b shows the gamma analysis for MLC position check and performance test, respectively. The mismatch in the gamma analysis was observed only in the peripheral regions. Pretreatment RapidArc quality assurance using 2D array and Octavius phantom Table 1 shows the result for the 10 different RapidArc plans. The results indicate that the measured value agrees well with that of the value calculated by TPS in the treated volume region. The percentage dose points failed the gamma criteria is 5%, except for the 2 head and neck cases, for which planning was done with double arc using the avoidance sector. Also the higher variation observed, particularly in the low dose regions outside the treatment volume, may be caused by complex MLC movements with small effective openings, which will make it more challenging to deliver the radiation. Discussion Characterizations of the response are very much essential for an ion chamber based detector before their clinical use. Ion chamber based detector arrays are known to have insignificant energy and dose rate dependence for megavoltage photon beams but require a large sensitive volume with a diameter of 5 mm for each chamber to gain a signal, and they will therefore exhibit a volume averaging effect in steep dose gradient regions. 18 2D array shows a very high response on short-term and longterm reproducibility, which should be essential for a dose measuring system. Also longer warm-up time seems to be needed to achieve more accurate results. 9 Studies indicate a slight increase of the 2D array readings as the field size is increased. This may be

58 S. A. Syamkumar et al. / Medical Dosimetry 37 (2012) 53-60 Fig. 9. Directional dependency for 6-MV and 15-MV arc deliveries. Gamma analysis results for cross-plane and in-plane profiles. Red color indicates the area where the gamma analysis fails. because the amount of low-energy contamination increases as the field size increases at a constant SSD. Furthermore, the mean photon energy of the primary radiation beam is reduced, with increasing field size as the proportion of the radiation incident on the phantom caused by scatter from the beam-defining system increases. 2D array exhibits a considerable percentage of directional dependency for static fields as the array is irradiated from the bottom side and parallel to the beam axis. So, in practice a correction factor for directional dependency should be included in the measurements. For all calculations in the TPS, the AAA algorithm has been used, which provides sufficient accuracy by accounting for the heterogeneity correction. 20,21 The advantage of air-filled detectors is their insensitivity to radiation damage, which is explained by Spezi et al. 17 Another advantage of air-filled detectors surrounded by materials of low atomic number is that they are free from any enhanced sensitivity to lowenergy scattered photons because of the photoelectric effect in materials of the array. The main limitations of the 2D array type 10024 are the geometrical resolution of the detector, the size of the single detector, and the center-to-center distance between the detectors. This efficiency problem is common to other planar detectors presently available. The main advantage of using the 2D array for RapidArc quality assurance measurements is the ability to perform absolute dose comparisons for hundreds of measurement positions using only a single-beam delivery, compared with the multiple, absolute point dose measurements to be done with a single ionization chamber. As our result shows, the 2D array seven29 detector can be used for pretreatment verification of RapidArc plans. Also it is important to normalize the dose distribution in the high-dose region and exclude the low-dose area. Quality assurance for the same patients was also done using other available phantoms with ion chambers and films. The direct Fig. 10. Dose profiles for 20 20-cm 2 open fields for 6 MV and different wedge-modulated profiles. (15, 30, 45, 60 ) for 6MV photon beams. Measured dose profiles compared with the semiflex ionization chamber measurements done in a water phantom at the same measurement setup. Profiles are normalized to the central axis beam.

S. A. Syamkumar et al. / Medical Dosimetry 37 (2012) 53-60 59 Fig. 11. (a). MLC quality assurance test using 2D array. Gamma analysis for split fields and chair pattern test for MLC positions. (b) Gamma analysis for sweep-field tests for MLC performance. comparison of gamma analysis with the other conventional methods was not meaningful because of the difference in the size and shape of the phantom. The other conventional methods were only used to demonstrate whether the RapidArc plans were calculated and delivered correctly. Also the measured and calculated profiles were compared qualitatively. To compensate for the anisotropic behavior of the 2D array during arc measurements with the Octavius phantom, a 2-cm compensation cavity has been provided in the bottom part for the Octavius linac phantom. This offers a better compensation for the directional dependency when the array is used for the RapidArc pretreatment patientspecific quality assurance. The reduced charge collection will be bal-

60 S. A. Syamkumar et al. / Medical Dosimetry 37 (2012) 53-60 Table 1 Gamma analysis results for pretreatment quality assurance of 10 RapidArc cases Sites anced by the removal of the adequate phantom material. The compensating cavity should extend up to the sides of the array. Because of practical difficulties, the cavity is limited only to the bottom side. We found from our studies that the 2D array with the Octavius phantom can be used for online verification of the RapidArc quality assurance, which can be done within a short period, when compared with the film analysis or Electronic Portal Imaging Device (EPID) measurements. The spatial resolution of film and EPID will be superior when compared with the 2D array measurements. 22,23 However, EPID requires correction factors for energy dependence. 24 The spatial resolution problem can be overcome in 2D array by merging the images in each of the measurements. Also film requires an expensive 2D film density scanner and software for converting optical density to dose and comparing the 2D dose distribution to the dose distribution provided by the TPS. 25 Setting, acquiring data, and analyzing will take about 30 minutes using the 2D array. Conclusion On the basis of the studies performed, it can be concluded that the 2D array seven29 has the necessary characteristics and can be used efficiently in a clinical setting. The 2D array provides an overall accuracy when compared with single ionization chamber measurements for static and rotational delivery. Moreover, the dose calibration for the 2D array is easy and stable. Also our studies have shown that 2D seven29 array with Octavius phantom is an efficient method for RapidArc patient-specific quality assurance with a satisfactory accuracy for clinical practice. References Patient Number Head and neck 1 4.72 2 2.85 3 5.88 4 7.11 Esophagus 5 1.39 6 1.31 Cervix 7 3.53 8 3.00 9 4.31 10 2.89 DTA distance to agreement, DD dose difference tolerance level Gamma Analysis for 3 mm DTA, 3% DD. % Points Failed 1. Webb, S. Intensity Modulated Radiation Therapy. Bristol, UK: Institute of Physics Publishing; 2000. 2. Yu, C.X. Intensity-modulated arc therapy with dynamic multileaf collimation: An alternative to TomoTherapy. Phys. Med. Biol. 40:1435 49; 1995. 3. Yu, C.X.; Li, X.A.; Ma, L.; et al. Clinical implementation of intensity-modulated arc therapy. Int. J. Radiat. Oncol. Biol. Phys. 53:453 63; 2002. 4. Mackie, T.R.; Holmes, T.; Swerdloff, S.; et al. TomoTherapy: A new concept for the delivery of dynamic conformal radiotherapy. Med. Phys. 20:1709 19; 1993. 5. Welsh, J.S.; Patel, R.R.; Ritter, M.A.; et al. Helical TomoTherapy: An innovative technology and approach to radiation therapy. Technol. Cancer Res. Treat. 1:311 6; 2002. 6. Poppe, B.; Blechschmidt, A.; Djouguela, A; et al. Two-dimensional ionization chamber arrays for IMRT plan verification. Med. Phys. 33:1005 15; 2006. 7. Jursinic, P.A.; Ben, E., Nelms, A. 2-D diode array and analysis software for verification of intensity modulated radiation therapy delivery. Med. Phys. 30:87 9; 2003. 8. Martens, C.; De Wagter, C., De Neve, W. The value of the LA48 linear ion chamber array for characterization of intensity-modulated beams. Phys. Med. Biol. 46:1131 48; 2001. 9. Jonathan, G.; Guanghua, L.; Yan, C.; et al. Comparison of two commercial detector arrays for IMRT quality assurance. J. Appl. Clin. Med. Phys. 10; 2009. 10. Poppe, B.; Blechschmidt, A.; Djouguela, A.; et al. Two-dimensional ionization chamber arrays for IMRT plan verification. Med. Phys. 33:1005 15; 2006. 11. Martens, C.; De Wagter, C., De Neve, W. The value of PinPoint ion chamber for characterization of small field segments used in intensity-modulated radiotherapy. Phys. Med. Biol. 45:2519 30; 2000. 12. Ling, C.C.; Zhang, P.; Archambault, Y.; et al. Commissionng and quality assurance of RapidArc radiotherapy delivery system. Radiat. Oncol. J. Biol. Phys. 72:575 81; 2008. 13. Van Esch, A.; Clermont, C.; Devillers, ML.; et al. On-line quality assurance of rotational radiotherapy treatment delivery by means of a 2D ion chamber array and the Octavius phantom. Med. Phys. 34:3825 37; 2007. 14. Earl, M.A.; Shepard, D.M.; Naqvi, S.; et al. Inverse planning for intensity-modulated arc therapy using direct aperture optimization. Phys. Med. Biol. 48:1075 89; 2003. 15. Shepard, D.M.; Cao, D.; Afghan, M.K.N.; et al. An arc-sequencing algorithm for intensity modulated arc therapy. Med. Phys. 34:464 70; 2007. 16. Low, D.A.; Harms, W.B.; Mutic, S.; et al. A technique for the quantitative evaluation of dose distributions. Med. Phys. 25:656 61; 1998. 17. Spezi, E.; Angeline, A.L.; Romani, F.; et al. Characterization of a 2D ion chamber array for the verification of radiotherapy treatments. Phys. Med. Biol. 50:3361 73; 2005. 18. Low, D.A.; Parikh, P.; Dempsey, J.F.; et al. Ionization chamber volume averaging effects in dynamic intensity modulated radiation therapy beams. Med. Phys. 30: 1706 11; 2003. 19. Laub, W.U., Wong, T. The volume effect of detectors in the dosimetry of small fields used in IMRT. Med. Phys. 30:341 7; 2003. 20. Van Esch, A.; Tillikainen, L.; Pyykkonen, J.; et al. Testing of the analytical anisotropic algorithm for photon dose calculation. Med. Phys. 33:4130 48; 2006. 21. Bragg, C.M., Conway, J. Dosimetric verification of the anisotropic analytical algorithm for radiotherapy treatment planning. Radiother. Oncol. 81:315 23; 2006. 22. Van Esch, A.; Depuydt, T., Huyskens, D.P. The use of a Si-based EPID for routine absolute dosimetric pre-treatment verification of dynamic IMRT fields. Radiol. Oncol. 71:223 34; 2004. 23. Van Esch, A.; Vanstraelen, B.; Verstraete, J.; et al. Pre treatment dosimetric verification by means of a liquid-filled electronic portal imaging device during dynamic delivery of intensity modulated treatment fields. Radiol. Oncol. 60:181 90; 2001. 24. Renner, W.D.; Norton, K., Holmes, T. A method for deconvolution of integrated electronic portal images to obtain incident fluence for dose reconstruction. J. Appl. Clin. Med. Phys. 6:22 9; 2005. 25. Wiezorek, T.; Banz, N.; Schwedas, M.; et al. Dosimetric quality assurance for intensity-modulated radiotherapy feasibility study for a filmless approach. Strahlenther. Onkol. 181:468 74; 2005.