|Year : 2017 | Volume
| Issue : 1 | Page : 39-44
Verification of an irregular field algorithm of a treatment planning system using a locally designed pelvic phantom: A simple design low-cost phantom suitable for quality assurance and control test in radiotherapy
Michael Onoriode Akpochafor, Akintayo Daniel Omojola, Samuel Olaolu Adeneye, Moses Adebayo Aweda, Muritala Shehu Oniyangi, Chieloka Chinazom Iloputaife
Department of Radiation Biology, Radiotherapy, Radiodiagnosis and Radiography, College of Medicine, Lagos University Teaching Hospital, Idi-Araba, Lagos, Nigeria
|Date of Web Publication||15-Feb-2017|
Akintayo Daniel Omojola
Department of Radiation Biology, Radiotherapy, Radiodiagnosis and Radiography, College of Medicine, Lagos University Teaching Hospital, Idi.-Araba, Lagos
Source of Support: None, Conflict of Interest: None
BACKGROUND: Modern radiotherapy treatment machine today comes along side with sophisticated treatment planning systems (TPSs). The accuracy of any TPS depends on the mathematical algorithm it uses and can be well verified using a dedicated phantom.
AIMS AND OBJECTIVES: To design a low-cost pelvic phantom and to use the designed phantom to verify whether the accuracy of an Irregular Field Algorithm of a Precise PLAN 2.16 TPS is within ±5% International Commission on Radiation Units and Measurements (ICRU) minimal limit.
MATERIALS AND METHODS: Designed pelvic phantom was made of Plexiglas with six tissue equivalent inserts and an ion-chamber port. The mimicked organs for the inserts were: Prostate, bladder, adipose, muscle, rectum, and bone. A Hi-speed computed tomography (CT) simulator was used for acquiring images and CT numbers of the designed pelvic phantom, a Precise PLAN 2.16 TPS was used for image planning, an Elekta-Precise Clinical Linear Accelerator was used for prescription of the planned images and a precalibrated NE 2570/1 farmer-type ion-chamber with an electrometer was used to calculate the mean dose. Data analysis value was done using GraphPad Prism 5.0 statistics software.
RESULTS: The maximum percentage deviation with large field sizes of 22 cm × 25 cm for six different inhomogeneous inserts was −3.95%, and bone only homogeneous inserts was 2.38%. The maximum percentage deviation with small field sizes of 5 cm × 5 cm with six different inhomogeneous inserts was −3.57%. The percentage deviation between the solid water phantom and the locally designed pelvic phantom was −3.46%.
CONCLUSION: The irregular field algorithm showed an overall accuracy of approximately ±4% with the locally designed pelvic phantom for both large and small field sizes against ±5% ICRU minimal limit. Although there were significant differences in percentage deviation between inhomogeneity and homogeneous insert irrespective of field sizes.
Keywords: Computed tomography, Hounsfield unit, irregular field algorithm, photon beam, Plexiglas, treatment planning system
|How to cite this article:|
Akpochafor MO, Omojola AD, Adeneye SO, Aweda MA, Oniyangi MS, Iloputaife CC. Verification of an irregular field algorithm of a treatment planning system using a locally designed pelvic phantom: A simple design low-cost phantom suitable for quality assurance and control test in radiotherapy. Int J Health Allied Sci 2017;6:39-44
|How to cite this URL:|
Akpochafor MO, Omojola AD, Adeneye SO, Aweda MA, Oniyangi MS, Iloputaife CC. Verification of an irregular field algorithm of a treatment planning system using a locally designed pelvic phantom: A simple design low-cost phantom suitable for quality assurance and control test in radiotherapy. Int J Health Allied Sci [serial online] 2017 [cited 2021 Dec 1];6:39-44. Available from: https://www.ijhas.in/text.asp?2017/6/1/39/200196
Radiotherapy is the treatment of disease, especially cancer cells, using X-rays, gamma rays electron, and proton beam to destroy or retard their growth. Radiation therapy has an increasingly important role in the medical field, particularly in the treatment of malignant diseases, such as cancer. Worldwide, about 40% of cancer patients require radiation treatment, either curative or palliative.,
The International Commission on Radiation Units and Measurements (ICRU) has recommended that radiation dose must be delivered to within ±5% of the prescribed dose.,, Although ICRU report 24 also states that recommended uncertainty in the delivered dose to a phantom (mimicked human subject) at “optimal model” should be ±2.5% and at “minimal” or “lowest acceptable” model be ±5% and International Atomic Energy Agency (IAEA) on-site dosimetry limit be ±3.0%.,, For a radiotherapy center using three-dimensional conformal radiotherapy (3D-CRT) treatment technique, there is the need to verify the accuracy of the irregular field algorithm in use which is based primarily on measured data, because quality assurance program ensures that all the components of the treatment facilities used in radiotherapy must be properly checked for accuracy and consistency and that all radiation generating facilities are functioning according to manufacturer's specification. Several technique of carrying out quality assurance of treatment planning system (TPS) has been proposed by various authors.,,,,, Likewise, the reduction of errors and uncertainties in the dose calculation plays an important role in the success of a treatment procedure. The performance and quality of any TPS is dependent on the type of algorithm used. Some algorithm in use includes: Irregular field algorithm (Precise PLAN), pencil beam convolution (Eclipse PBC), analytical anisotropic algorithm (Eclipse AAA), AcurosXB (Eclipse AXB), fast-Fourier Transform convolution (XiO Convolution), multigrid superposition (XiO Superposition), and Monte Carlo photon (Monaco MC).,,,,
An algorithm is defined as a sequence of instructions that operate on a set of input data, transforming that information into a set of output result that are of interest to the user. The intent of a dose calculation algorithm is to predict, with as much accuracy as possible, the dose delivered to any point within the patient for any given beam orientation. Precise PLAN 2.16 (Elekta, Crawley, UK) TPS, which enables 3D-CRT planning, was used for treatment plans. To calculate the dose distribution of the photon beam, the TPS uses an irregular field algorithm, for different depths and field sizes, based on data measures in a phantom. The algorithm takes into account the inhomogeneity of the patient's tissue and uses an integration scheme to evaluate the scatter component of the dose. The concept of this dosimetry of Irregular Fields using tissue maximum ratio and scatter-maximum ratios (SMRs) is analogous to the method using tissue-air ratio (TARs) and scatter-air ratios (SARs). The magnitude of the dose from scattered radiation at some given point can be quantified using the SARs, SMRs technique.,,
This study was conducted in the Department of Oncology at Lagos University Teaching Hospital in Nigeria where a Hi-speed computed tomography (CT) simulator, Precise PLAN 2.16 TPS, and an Elekta-Precise Clinical Linear Accelerator are in use. This study was therefore carried out to design a low-cost pelvic phantom and to use the phantom to verify the irregular field algorithm of the TPS if it meets ±5% ICRU “minimal” or “lowest acceptable” limit and to verify if the phantom can be used for routine quality assurance and control test in local radiotherapy centers. The irregular field algorithm which is based on Clarkson integration is given by the relation:
Where, TRAY and TRAY2 = Are the tray factors, Output = The output factor for 0 × 0 field size, FSC = The airfield size correction dependency factor, computed for equivalent square of the collimator opening, SSD = Source to surface distance, DMAX = Dose at maximum, SPD = Source to point distance of calculation, X and Y = Coordinates at depth of the point of calculation, c = Is the correction for virtual location of the source, QF = Quality factor, OCR = Off-center ratio, TAR0 = Tissue-air ratio at the surface.
| Materials and Methods|| |
The designed in-house pelvic phantom was made of Plexiglas of thickness 0.33 mm having a density 1.16 g/cm 3. A plastic based hardener (allplast) was used for holding one slab to another to form a cube. The Plexiglas used was purchased from a local plastic shop of dimension 4 by 8 feet, a part of which was cut using a plastic cutter into six slabs each of dimension 30 cm × 30 cm. Seven holes were drilled on one face. Each drilled hole had a diameter of 2.5 cm gummed together using plastic-based hardener called “allplast.” Before the holes were drilled, the distance from the surface of the locally designed pelvic phantom to the ion-chamber was 15 cm, while the distance between two diagonal inserts were approximately 22 cm. The distance from one insert to the other (horizontally) was 7 cm and vertically was 18 cm. Furthermore, additional drilled hole was made at the top of the locally designed pelvic phantom to allow for easy passage of water in and out of the phantom. After these holes and distance have been marked out, another cylindrical rod made of Plexiglas material of thickness 0.2 mm, length 14.3 cm, and diameter 2.5 cm were fitted into the seven drilled holes and were held together at the tip by the “allplast” gum to avoid leakage [Figure 1]. The phantom was loaded with tissue-equivalent material putting into consideration the attenuation coefficients, electron densities, and the effective atomic numbers of each chemical [Table 1]. The phantom was scanned under a Hi-speed CT simulator, and slices of images were acquired for six different tissue-equivalent materials and bone only equivalent material. The CT number (is also expressed in a standardized and convenient form as the Hounsfield unit [HU]) for the six different tissue equivalent materials was determined from the CT monitor [Table 2]. Images were transported through the Digital Imaging and Communications in Medicine (DICOM) to the Precise PLAN TPS where 12 fields technique denoted as Beam (BM) 1 –BM 12 were used with large field sizes covering the six inserts. The gantry angles were: 0°, 22.5°, 45°, 90°, 135°, 157.5°, 180°, 197.5°, 215°, 270°, 315°, and 337.5°, respectively. The total dose for the 12 fields was 100 cGy, and the total monitor unit (MU) was 100 HU. The type of beam used was “simple.” The photon energy used was 6 MeV, source to axis distance (SAD) was 100 cm and SSD was approximately 84 cm. Collimator angle was 0°, the diaphragm upper SAD was approximately 25 cm and lower SAD was 22 cm, giving at total area diaphragm size of 22 cm × 25 cm. Under modifiers, tray factor was 1 and no multileaf collimator (MLC) was present.
|Figure 1: Locally designed in-house pelvic phantom before computed tomography simulation|
Click here to view
|Table 1: Mimicked tissue equivalent materials and their chemical compositions|
Click here to view
|Table 2: Computed tomography number for different mimicked tissue equivalent material in Hounsfield unit compared with another study|
Click here to view
A second scan was carried using the same protocol using large field sizes, but the insert was only bone equivalent material. The acquired images from the CT simulator were also transferred to the Precise PLAN 2.16 TPS through a DICOM. Six fields technique were used denoted as BM 1–BM 6 covering the six inserts which were uniformly homogeneous. The gantry angles were: 0°, 45°, 90°, 180°, 225°, and 270° respectively The total dose for the 6 field was 100 cGy, and total MU was 100 MU. The type of beam used was “simple.” The photon energy used was 6 MeV; SAD was 100 cm and SSD was approximately 84 cm collimator angle was 0°, the diaphragm upper SAD was approximately 25 cm, and lower SAD was 22 cm, giving at total area diaphragm size of 22 cm × 25 cm. Under modifiers, tray factor was 1, and no MLC was present.
A third scan was done following the same protocol with small fields. Six fields technique denoted as BM 1–BM 6 were used covering the six inserts. The gantry angles were: 0°, 22.5°, 45°, 90°, 135°, 157.5°, 180°, 197.5°, 215°, 270°, 315°, and 337.5° respectively. The total dose for the six fields was 100 cGy and total MU making up 100 MU was prescribed. The type of beam used was “simple.” Photon energy of 6 MeV was used. SAD was 100 cm and SSD was approximately 84 cm collimator angle was 0°, the diaphragm upper SAD was approximately 5 cm and lower SAD was 5 cm, giving at total area diaphragm size of 5 cm × 5 cm. Under modifiers, tray factor was 1 and no MLC was present.
Finally, a simple experimental protocol was done to further verify the accuracy of the designed pelvic phantom and to verify the accuracy of the Precise PLAN 2.16 TPS Algorithm by comparing dose values (in Gy) between the locally designed pelvic phantom and solid water phantom (SWP) of the Elekta-Precise Linear Accelerator which was used as standard against the designed pelvic phantom with 100 cm SSD for both phantom. Six readings were also observed with the gantry angle at 0° (1 anterior).
The Elekta-Precise Clinical Linear Accelerator was calibrated using a large water phantom before this study was done using a 6 MeV photon beam to give 100 cGy (1 Gy) at 100 MU with a precalibrated NE 2570/1 farmer-type ion-chamber to determine the absorbed dose. Necessary corrections for temperature, pressure, polarization, recombination, etc., were effected on the ion-chamber response. Absorbed dose at reference depth was calculated as follows:
Where MQ is the electrometer reading (charge) corrected for temperature and pressure, ND, W is the chamber calibration factor and KQ, Q0 is the factor which corrects for difference in the response of the dosimeter at the calibration quality Q and at quality Q0 of the clinical X-ray beam according to the TRS-398 protocol of the IAEA. The deviation between the calculated and measured dose was obtained using the following equation:
Where Dc is calculated dose from the large water phantom and Dm measured dose result from the designed pelvic phantom for this study. The calculated value for Dc ≈ 100 cGy (1 Gy).
Data analysis value was done using GraphPad Prism 7.0 (GraphPad Software, Inc., San Diego, California, USA). Descriptive statistics and unpaired t-test with Welch's correction was implored at 95% level of significance. A P < 0.05 was considered statistically significant.
| Results|| |
A summary of the CT numbers also known as the HU for the mimicked chemical composition for prostate, bladder, adipose, muscle, rectum, and bone was 45 ± 5 HU, 47 ± 4 HU, –104 ± 10 HU, 50 ± 15 HU, 30 ± 17 HU, and –780 ± 30 HU, respectively [Table 2].
The mean doses, standard deviations, and percentage deviations for six different tissue-equivalent materials with twelve fields each of sizes 22 cm × 25 cm was determined with the with the minimum percentage deviation noticed in the third field (–0.17) and the maximum percentage deviation noticed in the ninth field (–3.95) [Table 3] and six fields each of sizes 5 cm × 5 cm with the minimum percentage deviation noticed in the first field (0.71) and the maximum percentage deviation noticed in the ninth field (–3.57) [Table 4].
|Table 3: Mean absorbed dose (Gy) and percentage deviation for pelvic inhomogeneous inserts with large field size (22 cm × 25 cm)|
Click here to view
|Table 4: Mean absorbed dose (Gy) and percentage of deviation for pelvic in-homogeneous inserts with small field size (5 cm × 5 cm)|
Click here to view
In addition, the mean doses, standard deviations, and percentage deviations for bone only homogeneous equivalent material with six fields with field size of 22 cm × 25 cm was as well determined with the minimum percentage deviation noticed in the sixth field (–0.054) and the maximum percentage deviation noticed in the third field (2.38) [Table 5].
|Table 5: Mean absorbed dose (Gy) and percentage deviation for pelvic bone only homogeneous inserts with large field size (22 cm × 25 cm)|
Click here to view
The comparison was made between the mean dose (Gy) of the locally designed pelvic phantom and SWP directly from the linear accelerator. Their mean doses were 0.6065 and 0.6275 Gy respectively, and percentage deviation that exists between them was 3.46% [Table 6].
|Table 6: Percentage deviation between solid water phantom against in-house pelvic phantom at gantry angle of 0°|
Click here to view
| Discussion|| |
A study has been carried out to assess the designed pelvic phantom and to verify the performance of a TPS which uses an irregular field algorithm based on previously published methods.,,,
The CT numbers from this study were compared to Schaly et al. CT numbers in a study to evaluation an anthropomorphic male pelvic phantom for image-guided radiotherapy. There was no significant difference in Schaly et al. CT numbers and this study's CT numbers (P > 0.05) showing that CT numbers obtained for this study was still within acceptable range.
The results were within the range of ±5% as recommended by ICRU  and were consistent with Van Dyk whose variation was within ±4%. Brahme  and Mijnheer et al. were within 3%–3.5% against this study which was approximately within 2.4%–4%. A study by Stathakis et al. using a 6-MV Philips SL 75-5 linear accelerator, show that the percentage deviation of measured and calculated dose was within 1% for large and small field sizes for inhomogeneity correction algorithm for irregular fields. It was seen to be more accurate than those obtained for this study whose for large and small field sizes for inhomogeneity correction was 3.95%.
Percentage deviation for pelvic inhomogeneous insert for large field size and small field sizes were above optimal limit (>±2.5%) but less than the minimum limit (<±5%). The percentage deviation for bone only pelvic homogeneous insert for large field size was <±2.5%. Pointing that, better accuracy was observed with homogeneous materials (that is bone only). Although inhomogeneity (that is insert with six different tissue equivalent materials) was still <±5%.
The irregular field algorithm had a maximum percentage deviation of –3.95% and + 2.38% for six inhomogeneous inserts and bone only homogeneous inserts with a large field size of 22 cm × 25 cm respectively [Table 3] and [Table 5]. The result for all six inhomogeneous inserts with small field size of 5 cm × 5 cm had a maximum percentage deviation of –3.57% [Table 4].
In addition, the result showed that the percentage deviation between the SWP and the locally designed pelvic phantom was –3.46%. The reason for this high deviation was because the linear accelerator SWP was homogeneous while that of the phantom which had six different tissue equivalent materials was inhomogeneous, thereby contributing to the increase of the percentage deviation. Although its value was less than the maximum percentage deviation for those obtained for large field size and small field size for inhomogeneous inserts. The value obtained was within accepted range, and it further confirms the accuracy of the designed pelvic phantom [Table 6].
An increase in the numerical value for the first six fields was noticed, with the highest deviation noticed in the tenth field. The overall percentage deviation range was –0.170%––3.95% [Table 3].
The least percentage deviation was noticed in the sixth field, and the highest deviation noticed in the third with an overall percentage deviation range of –0.054–+2.38%. In addition, it was observed that sixth field for bone only homogeneous had the least deviation of –0.054% [Table 5].
There was statistically significant difference between the percentage deviation for six tissue equivalent materials (prostate, bladder, adipose, muscle, rectum, and average bone) for inhomogeneous insert with large field size of 22 cm × 25 cm and 1 tissue equivalent material (bone only) homogeneous insert with large field size of 22 cm × 25 cm (P = 0.047). It was observed that independent of the field size, the six inhomogeneous had higher percentage deviation than bone only homogeneous insert.
There was no statistically significant difference between percentage deviations for large field size 22 cm × 25 cm for six equivalent tissue material against smaller field size 5 cm × 5 cm (P = 0.468). This result shows that the field sizes do not really have an effect on the outcome of the mean dose (Gy).
Furthermore, there was statistically significant difference between percentage deviation for six equivalent inhomogeneous tissue materials and bone only homogeneous material for small field size of 5 cm × 5 cm (P = 0.0197). The result also shows that inhomogeneous involving different tissue equivalent material tend to have more deviation while materials that are homogeneous have lesser deviation as seen in bone only irrespective of their field size.
The overall results show that the designed in-house phantom was approximately ±4% pointing that the irregular field algorithm compensated for both small and large field sizes. The clinical implication is that when treating a region of the body with different tissue density like the pelvic, the irregular field algorithm will adequately compensate for inhomogeneity and the error that may arise will be within clinically accepted range which ideal for any treatment machine in radiotherapy. According to the result obtained, the irregular field algorithm for homogeneity will further have reduced error due to similar tissue density.
| Conclusion|| |
A low cost locally designed phantom has been made for use in radiotherapy centers in Nigeria with about six centers using the same Precise PLAN 2.16 TPS. The designed in-house phantom showed an overall accuracy of ±4% with the irregular field algorithm of the TPS which falls within the acceptable range of ±5% set by ICRU. The designed in-house phantom materials were inexpensive and can be easily gotten from local chemical shops. The phantom proves highly useful for quality assurance and control test.
We would like to thank the Department of Radiation Biology and Radiotherapy of Lagos University Teaching Hospital.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Donaldson S. Towards Safer Radiotherapy. London, UK: British Institute of Radiology, Institute of Physics and Engineering in Medicine, National Patient Safety Agency, Society and College of Radiographers, and The Royal College of Radiologists; 2007.
Frödin JE, Jonsson E, Möller T, Werkö L. Radiotherapy in Sweden – A study of present use in relation to the literature and an estimate of future trends. Acta Oncol 1996;35:967-79.
Absorbed Dose Determination in External Beam Radiotherapy. IAEA Technical Reports Series. No. 398. Vienna, Austria: IAEA Publications; 2000.
Alam R, Ibbott GS, Pourang R, Nath R. Application of AAPM radiation therapy committee task group 23 test package for comparison of two treatment planning systems for photon external beam radiotherapy. Med Phys 1997;24:2043-54.
Animesh. Advantages of multiple algorithm support in treatment planning system for external beam dose calculations. J Cancer Res Ther 2005;1:12-20.
International Commission on Radiation Units and Measurements. Determination of Absorbed Dose in Patient Irradiated by Means of X or Gamma Rays in Radiotherapy Procedures. ICRU Report 24. Bethesda: ICRU Publications; 1977.
Izewska J, Dutreix A, Followill DS, Nisbet A, Novotny J, Sipila P, et al
. Standardized Quality Audit Procedures for On-site Dosimetry Visits to Radiotherapy Hospitals. Report of the IAEA Consultants' Meeting, IAEA, Vienna, 27 September-1 October, 1999; Revised in 2001. SSDL Newsletter 46. Vienna: IAEA; 2002.
International Atomic Energy Agency. On-site Visits to Radiotherapy Centres. Medical Physics Procedures. Vienna: IAEA Publications; 2007.
International Atomic Energy Agency. Setting Up a Radiotherapy Programme: Clinical, Medical Physics, Radiation Protection and Safety Aspects. Vienna: IAEA Publications; 2008.
Brahme A. Dosimetric precision requirements in radiation therapy. Acta Radiol Oncol 1984;23:379-91.
Clarkson JR. Note on depth doses in fields of irregular shapes. Brit J Radiol 1941;14:265-8.
Cygler J, Ross J. Electron dose distributions in an anthropomorphic phantom – Verification of theraplan treatment planning algorithm. Med Dosim 1988;13:155-8.
Fraass BA. Quality assurance for 3-D treatment planning. In: Palta J, Mackie TR, editors. Teletherapy: Present and Future. Madison, WI: Advanced Medical Publishing; 1996. p. 253-318.
Fraass B, Doppke K, Hunt M, Kutcher G, Starkschall G, Stern R, et al.
American Association of Physicists in Medicine Radiation Therapy Committee Task Group 53: Quality assurance for clinical radiotherapy treatment planning. Med Phys 1998;25:1773-829.
Institute of Physics and Engineering in Medicine (IPEM). Physics Aspects of Quality Control in Radiotherapy. IPEM Report 81. York: IPEM; 1999.
International Commission on Radiation Units and Measurements (ICRU). Dose Specifications for Reporting External Beam Therapy with Photons and Electrons. ICRU Report 29. Baltimore: Bethesda; 1978.
Jacky J, White CP. Testing a 3-D radiation therapy planning program. Int J Radiat Oncol Biol Phys 1990;18:253-61.
Johns HE, Cunningham JR. The Physics of Radiology. 3rd
ed. Springfield, IL: Charles C. Thomas; 1971. p. 362-3.
Khan FM. The Physics of Radiation Therapy. Baltimore: Lippincott Williams and Wilkins; 1984. p. 787-94.
Tan YI, Metwaly M, Glegg M, Baggarley S, Elliott A. Evaluation of six TPS algorithms in computing entrance and exit doses. J Appl Clin Med Phys 2014;15:4739.
Andic F, Baz Cifci S, Ors Y, Niang U, Dirier A, Adli M. A dosimetric comparison of different treatment plans of palliative spinal bone irradiation: Analysis of dose coverage with respect to ICRU 50 Report. J Exp Clin Cancer Res 2009;28:2.
Khan FM. The Physics of Radiation Therapy. Baltimore: Lippincott, Williams and Wilkins; 2003.
Van Dyk J. A Compendium for Medical Physicists and Radiation Oncologists. Madison, Wisconsin: Medical Physics Publishing; 1999. p. 235-8.
Cunningham JR. Scatter-air ratios. Phys Med Biol 1972;17:42-51.
Schaly B, Varchena V, Au P, Pang G. Evaluation of an anthropomorphic male pelvic phantom for image-guided radiotherapy. Rep Med Imaging 2009;2:69-78.
Mijnheer BJ, Battermann JJ, Wambersie A. What degree of accuracy is required and can be achieved in photon and neutron therapy? Radiother Oncol 1987;8:237-52.
Podgorsak EB. Radiation oncology physics. A Handbook for Teachers and Students. Vienna, Austria: IAEA Publication; 2005.
Kappas C, Rosenwald JC. Quality control of inhomogeneity correction algorithms used in treatment planning systems. Int J Radiat Oncol Biol Phys 1995;32:847-58.
Van Dyk J, Barnett RB, Cygler JE, Shragge PC. Commissioning and quality assurance of treatment planning computers. Int J Radiat Oncol Biol Phys 1993;26:261-73.
Stathakis S, Kappas C, Theodorou K, Papanikolaou N, Rosenwald JC. An inhomogeneity correction algorithm for irregular fields of high-energy photon beams based on Clarkson integration and the 3D beam subtraction method. J Appl Clin Med Phys 2006;7:1-13.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]