About this Journal Submit a Manuscript Table of Contents
BioMed Research International
Volume 2013 (2013), Article ID 436979, 8 pages
http://dx.doi.org/10.1155/2013/436979
Research Article

Evaluation of Wall Correction Factor of INER's Air-Kerma Primary Standard Chamber and Dose Variation by Source Displacement for HDR 192Ir Brachytherapy

1Health Physics Division, Institute of Nuclear Energy Research, Longtan 325, Taiwan
2Department of Radiation Oncology, China Medical University Hospital, Taichung 404, Taiwan
3Department of Biomedical Imaging and Radiological Sciences, National Yang-Ming University, Taipei 112, Taiwan

Received 12 May 2013; Accepted 28 May 2013

Academic Editor: Maria F. Chan

Copyright © 2013 J. H. Lee et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

The aim of the present study was to estimate the wall effect of the self-made spherical graphite-walled cavity chamber with the Monte Carlo method for establishing the air-kerma primary standard of high-dose-rate (HDR) 192Ir brachytherapy sources at the Institute of Nuclear Energy Research (INER, Taiwan). The Monte Carlo method established in this paper was also employed to respectively simulate wall correction factors of the 192Ir air-kerma standard chambers used at the National Institute of Standards and Technology (NIST, USA) and the National Physical Laboratory (NPL, UK) for comparisons and verification. The chamber wall correction calculation results will be incorporated into INER's HDR 192Ir primary standard in the future. For the brachytherapy treatment in the esophagus or in the bronchi, the position of the isotope may have displacement in the cavity. Thus the delivered dose would differ from the prescribed dose in the treatment plan. We also tried assessing dose distribution due to the position displacement of HDR 192Ir brachytherapy source in a phantom with a central cavity by the Monte Carlo method. The calculated results could offer a clinical reference for the brachytherapy within the human organs with cavity.

1. Introduction

In radiotherapy treatments for cancer patients it is critical to have an accurate measurement of the dose delivered to the patient. Obtaining an accurate dose involves three major steps. The first step is the establishment of primary standards of air kerma or absorbed dose to water. The second step is the use of dosimetry protocols based on ion chambers calibrated using these primary standards to establish the dose under reference conditions in a clinical therapy beam. The final step is to establish the dose distribution in individual patients specified by computed tomography (CT) data.

At present, 192Ir is the most commonly used radioisotope for high-dose-rate (HDR) brachytherapy treatment. Due to the relatively short half-life of 192Ir (73.827 days ± 0.013 days) [1], most radiotherapy departments change their 192Ir sources every three months and medical physicists need to measure the source strength of the 192Ir sources on a regular basis before an accurate treatment plan can be written. The source calibration is a main component of quality assurance programs recommended for HDR brachytherapy [2]. The recommended quantity for the specification of brachytherapy gamma ray sources is the reference air-kerma rate (RAKR), defined by the International Commission on Radiation Units and Measurements [3, 4] as the kerma rate to air, in air, at a reference distance of 1 meter, corrected for air attenuation and scattering.

To replace the interpolation techniques between air-kerma calibration coefficients of an ionization chamber for 192Ir [5, 6], the Institute of Nuclear Energy Research (INER, Taiwan) has recently developed a spherical graphite-walled cavity ionization chamber as the primary standard for direct measurement of HDR 192Ir brachytherapy source strength to provide a traceable source calibration. Our primary standard is derived from measurements using the graphite-walled chamber, based on the Bragg-Gray theory. The wall correction factor is intended to account for the effects of attenuation and scatter of the incident primary photons in the chamber wall [7]. The of the primary standard chamber for HDR 192Ir brachytherapy sources is the largest correction factor (approximately 80% of the total correction amounts) required to correct the measured charge for experimental perturbations. In the past, the empirical method to estimate has been to measure the ionization charge (or current) as a function of wall thickness for a fixed cavity size (but for wall thickness no smaller than the minimum required to exclude secondary electrons generated outside the wall). The results are then linearly extrapolated to zero wall thickness under the assumption that attenuation and scattering are thus eliminated. For more than a decade, Rogers and Bielajew suggested that the use of based on linear extrapolation measurements was incorrect and proposed instead the use of results from the Monte Carlo calculations [812].

In this research, one of the tasks was to evaluate of the self-fabricated chamber by the Monte Carlo photon-electron transport calculations to establish the HDR 192Ir air-kerma primary standard at INER. In clinical brachytherapy treatment of the esophagus or bronchi or such organs with a cavity, the source positions tend to displace because there would be no body tissue to fix the isotope. The fact is that the position of the isotope would affect the accuracy of the output dose. So the source displacement would cause the tumor or normal tissue to receive inaccurate dose and deviate from the treatment plan and desired effect. Since the experiment of actually measuring the dose variations in the human body is not feasible, we used a simulation calculation to perform dose evaluation [1315]. In this research, we also performed sensitivity assessment using the Monte Carlo method for the displacement of an HDR 192Ir source in the cavity and explored how the related effects would affect internal doses. Hopefully the evaluation results will offer a clinical reference for brachytherapy within human organs with a cavity.

2. Materials and Methods

2.1. Wall Correction Factor Calculation for HDR 192Ir Air-Kerma Standard Chamber

The INER uses a Nucletron microSelectron HDR Classic brachytherapy unit fitted with the “Classic” source, part number 096.001, manufactured by Mallinckrodt Medical B V (The Netherlands). The average photon energy for HDR 192Ir brachytherapy sources is close to 0.4 MeV [16]. Figure 1 shows a schematic diagram of the HDR 192Ir brachytherapy source simulated in this work. The enclosure of the radioactive material consists of a cylindrical stainless steel AISI 316L capsule (length: 5.0 mm, radial thickness: 250 μm) which is sealed by laser welding. The 192Ir is contained in the capsule as a metallic 192Ir cylinder (length: 3.5 mm, diameter: 0.6 mm). The stainless steel capsule is welded to a metal plug and a 1500 mm long flexible stainless steel AISI 316 cable. The other end of the capsule is welded to a steel pin (tail). The identification of the source is engraved on the long side of the tail. The nominal initial activity of the source is between 370 GBq and 550 GBq.

436979.fig.001
Figure 1: Geometric model of the Nucletron mircoSelectron HDR 192Ir Classic source.

The INER primary standard cavity chamber for HDR 192Ir sources, shown in Figure 2, is a guarded ionization chamber resulting in low leakage currents. The spherical cavity volume of the primary standard chamber was measured on two coordinate measuring machines and found to be 102 cm3. The outside radius (3.200 cm), inside radius (2.899 cm), and wall thickness (0.301 cm) for the chamber were measured with a similar technique. In the case of 192Ir, this requires a wall thick enough to stop 687 keV Compton recoil electrons generated by 885 keV gamma rays, the most energetic photons emitted by 192Ir [17], neglecting three very weak lines above 1 MeV. The CSDA (continuous slowing down approximation) range of 687 keV electrons is 0.31 g cm−2 of graphite [18], which is equivalent to a wall thickness of approximately 1.8 mm. The graphite wall of this cavity chamber provides sufficient build-up material to ensure CPE. The computer code CAVSPHnrc, which is a user code of EGSnrc [19], is used to calculate wall correction factors of standard cavity chambers. The ionization chamber used in the calculation comprised two concentric spheres, and if ignoring the central electrode, there would be three areas including air, graphite wall, and cavity (made of air) from the outside to the inside. It was assumed that the 192Ir source was a point source and that the distance between the source and ionization chamber was 100 cm. The simulation kept track of the wall from which the electrons entered the cavity and calculated and , the correction factors for attenuation and scattering in the wall. From Figure 3 the following equations become evident: where the primary photon collides with the chamber wall and produces electrons and a scattered photon, is the energy deposition caused by the primary photon, is the energy deposition caused by the scattered photon, and which stands for all the energy deposition inside the chamber cavity. If the ionization chamber was wall-less, the energy deposition in the cavity would be the energy deposition of the primary photon before attenuation, that is, . Hence, the definition of the wall correction factor, , would be the ratio of the actual energy deposition to the wall-less energy deposition . In order to produce results that could be used for arbitrary beam spectra, calculations for the self-made chamber were done for both monoenergetic and spectral photon beams. Samples of at least 108 incident primary photons were used to make the relative statistical standard deviations below 0.1%.

436979.fig.002
Figure 2: Schematic of INER primary standard cavity chamber for HDR 192Ir source.
436979.fig.003
Figure 3: The Monte Carlo simulation geometry for wall correction assessment of the INER primary standard cavity chamber for HDR 192Ir source.
2.2. Assessment of Dose Variation by Source Displacement for HDR 192Ir Brachytherapy

Plexiglass is a useful phantom material for experimental measurement and calibration and can be easily custom-made for certain shapes and purposes. The radiation interaction properties and density of Plexiglass are similar to the human soft tissue. The Plexiglass phantom is a cylinder with a height of 14 cm and a radius of 10 cm, as indicated in Figure 4. In the center of the phantom, there is a 2 cm diameter cylindrical cavity, which in this calculation represents human organs with a cavity such as the esophagus and bronchi. At the edge of the Plexiglass phantom there are also four cylindrical cavities of the same size as that of the central one. In this calculation the cavities at the edge were not simulated and were filled with the phantom material. To simulate the movement of the 192Ir HDR source in the cavity, the source was located at the cylindrical cavity surface in the calculation, at point A and point B in Figure 5. The HDR 192Ir source was placed at a height of 7 cm in the cavity, which is the middle of the cavity length. As Figure 5 showed, the distance between point A and B is exactly the diameter of the cylindrical cavity. The photon and electron radiations were simulated separately with a Nucletron microSelectron Classic source with an activity of 3.552 × 1011Bq (9.6 Ci) and then the doses due to photon and electron radiations were summed. The geometrical model of the dose monitoring positions is also indicated in Figure 5, taking the cube whose side length was 0.5 cm as a scoring volume. The monitoring positions went along as a straight line from the radius of the central cavity to the side surface of Plexiglass phantom. In this work, the doses at the same monitoring points while the 192Ir source was positioned at the central axis of the cavity were also calculated as the comparison basis for the impact of the source displacement.

436979.fig.004
Figure 4: Diagram of Plexiglass phantom geometry.
436979.fig.005
Figure 5: Geometrical indication for the positions of HDR 192Ir source and dose assessment scoring volume in Plexiglass phantom.

3. Results and Discussions

3.1. Calculation and Verification of the Chamber Wall Correction Factors

RAKR is determined in air at a distance of 1 m from the source axis in the plane that perpendicularly bisects the axis. The photon spectrum at the RAKR measurement point was estimated by assuming the photon emission probabilities for 192Ir decay and calculating the attenuated spectrum reaching the measurement point. The calculation took into account the attenuation along all photon paths through the various materials by integrating overall source points in the cylindrical core. The photon intensity and spectrum for the source at the RAKR determination point was calculated using MCNP code version 5 [20]. In the calculation the geometry model of the HDR 192Ir source was constructed and simulated in detail. The energy cut-off settings for the photon and electron transport were both 0.001 MeV. Photons were sampled uniformly in the core and were scored at the side of the HDR 192Ir source. Figure 6 shows the 192Ir source photon spectrum calculated in this research, with which the spectrum for the similar field from the independent Monte Carlo calculations of Rogers and Borg [21] showed a good agreement.

436979.fig.006
Figure 6: Fluence spectrum for the Nucletron mircoSelectron HDR 192Ir Classic source at the RAKR measurement point.

The 192Ir source photon spectrum evaluated in Figure 6 was taken as the input data of the CAVSPHnrc code and was used to calculate the wall correction factors of INER’s spherical graphite-wall cavity chamber. The , , and calculation values and simulation uncertainty analysis for INER’s HDR 192Ir primary standard chamber are listed in Table 1. Figure 7(a) illustrates the wall correction factors and the related parameters of INER’s spherical graphite-wall cavity chamber, , , and , as functions of the incident photon energy. The value reduced with the increase of energy. When compared with the mass attenuation coefficient () of graphite, it showed similar variation trends (Figure 7(b)) and met the proportional relationship between and in (2). varied inversely with the incoherent scattering cross section (Figure 7(c)) related to the energy deposition caused by the scattered photon in (3). The smallest value occurred at 0.04 MeV. The wall correction factors affected by the related parameters have been established in this study and will be incorporated into INER’s air-kerma primary standard for HDR 192Ir sources.

tab1
Table 1: The , , and calculation values and simulation uncertainty analysis for INER’s HDR 192Ir primary standard chamber.
fig7
Figure 7: (a) Wall correction factors , attenuation correction factor , and scatter correction factor , (b) and mass attenuation coefficient of graphite , and (c) and incoherent scattering cross section as functions of the incident photon energy.

INER used the CAVSPHnrc code to calculate the values of NIST’s spherical graphite primary standard chambers of different volumes for 0.40 MeV, 0.662 MeV, and 1.25 MeV photons and compared them with the results evaluated by NIST [22]. The comparison results are listed in Table 2. Table 2 shows that the assessment discrepancies against the NIST primary standard chambers between INER and NIST for monoenergetic photons were within 0.20%, verifying the accuracy and reliability of INER’s method for evaluating the chamber wall correction factors. The Monte Carlo simulation process in this study also offered a convenient and accurate approach to evaluate the chamber wall correction factor for national metrology institutes (NMIs) in establishing their air-kerma primary standards of gamma radiation. Table 3 gives the structural information of 192Ir primary standard cavity chambers for INER, NIST (50cc-1), and NPL [1, 22]. INER used the Nucletron microSelectron Classic source spectrum simulated in Figure 6 and the CAVSPHnrc code to calculate NIST’s and NPL’s 192Ir primary standard chamber wall correction factors given in Table 4. It can be seen from Table 4 that for the assessment of wall correction factors, the difference between evaluations of NPL’s standard chamber was 0.23% and between evaluations of NIST’s standard chamber was 0.44%. The differences were analyzed and the root cause was found: the HDR 192Ir source used by the NPL was the Nucletron microSelectron Classic type [16], which was the same type used by INER for calculation of the HDR 192Ir source spectrum. On the other hand, NIST used a low-dose-rate (LDR) 192Ir brachytherapy seed source to evaluate the wall correction factor [22]. With the different types of sources, INER and NIST had a larger difference in the evaluation results. Analyzing Tables 2 and 4, it can be seen that for the NIST 50cc-1 primary standard chamber for an 192Ir source, the evaluation difference of the 192Ir spectral photon beams is higher than that for the monoenergetic photons (0.40 MeV, 0.662 MeV, and 1.25 MeV).

tab2
Table 2: Comparison of values calculated by INER and NIST for NIST’s spherical graphite primary standard chambers using 0.40 MeV, 0.662 MeV, and 1.25 MeV photons.
tab3
Table 3: Structural characteristics of 192Ir primary standard cavity chambers for INER, NIST (50cc-1), and NPL.
tab4
Table 4: Calculation comparison for wall correction factors of the 192Ir primary standard chambers for NPL and NIST.
3.2. Assessment of Dose Variation by Source Displacement

The MCNP code version 5 was adopted to evaluate the influence on the dose distribution of the phantom for HDR 192Ir brachytherapy source movement in the central cavity of the Plexiglass phantom. Figure 8 indicates the photon dose, electron dose, and the total dose as a function of distance when the source was placed on the central axis of the central cylindrical cavity. It is obvious that the dominant dose contribution came from the photons of the source. At the distance of 1.25 cm from the central cavity axis, the electron dose from the source was about 1/7 of the photon dose. At other distance monitoring points, the dose contribution of the electrons was only about 1/60 of the photon dose. The difference of these compared results as a function of distance was caused by the short range of the electrons. The electrons deposit energy quickly as they transport through 0.5 cm depth of the phantom, and the electron dose at the phantom edge results from the Bremsstrahlung. The dose contribution from the Bremsstrahlung was estimated to be approximately 17% of the total electron dose summed from all monitoring points. The dose assessment in the Plexiglass phantom showed that the total dose on the surface of the phantom was as low as only 1% of that at the phantom center.

436979.fig.008
Figure 8: Indication of the dose distribution varying with assessment distance while HDR 192Ir source was placed at the central axis of the Plexiglass phantom cylindrical cavity.

Figure 9 indicates that the total dose distribution varied with assessment distance when the 192Ir source was placed at different locations in the cylindrical cavity of the Plexiglass phantom. The dominant dose contribution is still from the photons, even if the location of the 192Ir source has been displaced. Table 5 lists the dose ratios at various monitoring points when the 192Ir source was placed at the point A, point B, and the central axis of the phantom cavity center. It was seen that for the nearest monitoring point (1.25 cm), the dose ratios of which the source was located at point A and point B against the source located at the cavity central axis were 0.340 and 16.2, respectively. The analysis results showed that when the location of the 192Ir source was displaced from point A to point B, the dose at the nearest motoring point (1.25 cm) was greater by nearly a factor of 50. The dose evaluation difference from the location of the 192Ir source decreased as the monitoring distance increased. At the edge of the Plexiglass phantom (9.75 cm), the dose difference was reduced to 28.5%. From the above calculation results, it can be known that when performing brachytherapy for organs with a cavity, the displacement of source position could make the dose output very different from the plan. With the increase of the assessment distance, the impact from the source displacement would be greatly reduced. However, for clinical brachytherapy, the source is usually very close to the tumor. Source position displacement in organs with cavities such as the esophagus and the bronchi may cause the delivered dose to differ from the prescribed dose in the treatment plan and may bring unexpected influence to the treatment results.

tab5
Table 5: Total dose distribution assessments when 192Ir source was at different locations.
436979.fig.009
Figure 9: Indication of the total dose distribution varying with assessment distance while the HDR 192Ir source was placed at different locations in the Plexiglass phantom cylindrical cavity.

4. Conclusions

We calculated for a self-made spherical chamber using the Monte Carlo method. The wall correction factor could be applied in the establishment of an air-kerma primary standard for HDR 192Ir brachytherapy sources. Simulation comparisons for the primary standard chamber wall corrections of different laboratories were employed to verify the accuracy of the assessment approach established by INER. The comparison results showed that INER and NPL had a better agreement with wall correction factor evaluation of NPL’s standard chamber using the same type of source as compared with the different type of 192Ir source used with NIST’s standard chamber, which resulted in an increased discrepancy. The calculation data in this study will be incorporated into INER’s air-kerma primary standard for HDR 192Ir sources in the near future and increase the calibration accuracy in brachytherapy source strength measurements.

According to the comparison results of the dose distribution of the Plexiglass phantom for the HDR 192Ir brachytherapy source, the outcome of brachytherapy will be seriously influenced by the source displacement. When dealing with clinical treatments including an internal cavity such as esophagus or bronchi, doctors and medical physicists should pay special attention to the dose output variations caused by the source that could not be exactly fixed. The MCNP simulation model used in this research hopefully could be applied to the Voxel phantom which is constructed by CT images to offer more contributions to improve clinical patient treatments.

Acknowledgments

The authors would like to thank the staff members at the Radiation Safety Assessment Group of the Institute of Nuclear Energy Research for their assistance in this study. They also extend their thanks to Dr. Lih-Yih Liao, Mr. Wei-Han Chu and Ms. Jui-Mei Deng for their useful comments and discussions. This study was supported by Grants CMU100-S-28 and DMR-99-107 from China Medical University Hospital.

References

  1. T. Sander and R. F. Nutbrown, “The NPL air kerma primary standard TH100C for high dose rate 192Ir brachytherapy sources,” Tech. Rep. DQL-RD 004, Teddington National Physical Laboratory, 2006.
  2. D. Baltas, K. Geramani, G. T. Ioannidis et al., “Comparison of calibration procedures for 192Ir high-dose rate brachytherapy sources,” International Journal of Radiation Oncology Biology Physics, vol. 43, no. 3, pp. 653–661, 1999. View at Publisher · View at Google Scholar · View at Scopus
  3. ICRU, “Dose and volume specification for reporting intracavitary therapy in gynecology,” ICRU Report 38, 1985.
  4. ICRU, “Dose and volume specification for reporting interstitial therapy,” ICRU Report 58, 1997.
  5. L. Büermann, H.-M. Kramer, H. Schrader, and H.-J. Selbach, “Activity determination of 192Ir solid sources by ionization chamber measurements using calculated corrections for self-absorption,” Nuclear Instruments and Methods in Physics Research A, vol. 339, no. 1-2, pp. 369–376, 1994. View at Scopus
  6. E. van Dijk, I.-K. Kolkman-Deurloo, and P. M. G. Damen, “Determination of the reference air kerma rate for 192Ir brachytherapy sources and the related uncertainty,” Medical Physics, vol. 31, no. 10, pp. 2826–2833, 2004. View at Publisher · View at Google Scholar · View at Scopus
  7. A. Piermattei, L. Azario, A. Fidanzio et al., “The wall correction factor for a spherical ionization chamber used in brachytherapy source calibration,” Physics in Medicine and Biology, vol. 48, no. 24, pp. 4091–4103, 2003. View at Publisher · View at Google Scholar · View at Scopus
  8. D. W. O. Rogers, A. F. Bielajew, and A. E. Nahum, “Ion chamber response and Awall correction factors in a 60Co beam by Monte Carlo simulation,” Physics in Medicine and Biology, vol. 30, no. 5, pp. 429–443, 1985. View at Publisher · View at Google Scholar · View at Scopus
  9. A. F. Bielajew, “Ionisation cavity theory: a formal derivation of perturbation factors for thick-walled ion chambers in photon beams,” Physics in Medicine and Biology, vol. 31, no. 2, pp. 161–170, 1986. View at Publisher · View at Google Scholar · View at Scopus
  10. A. F. Bielajew, “On the technique of extrapolation to obtain wall correction factors for ion chambers irradiated by photon beams,” Medical Physics, vol. 17, no. 4, pp. 583–587, 1990. View at Publisher · View at Google Scholar · View at Scopus
  11. D. W. O. Rogers and A. F. Bielajew, “Wall attenuation and scatter corrections for ion chambers: Measurements versus calculations,” Physics in Medicine and Biology, vol. 35, no. 8, pp. 1065–1078, 1990. View at Publisher · View at Google Scholar · View at Scopus
  12. D. W. O. Rogers and J. Treurniet, “Monte Carlo calculated wall and axial non-uniformity corrections for primary standards,” Tech. Rep. PIRS-633, National Research Council of Canada, Ottawa, Canada, 1999.
  13. A. Angelopoulos, P. Baras, L. Sakelliou, P. Karaiskos, and P. Sandilos, “Monte Carlo dosimetry of a new 192Ir high dose rate brachytherapy source,” Medical Physics, vol. 27, no. 11, pp. 2521–2527, 2000. View at Publisher · View at Google Scholar · View at Scopus
  14. F. Ballester, J. Pérez-Calatayud, V. Puchades et al., “Monte Carlo dosimetry of the Buchler high dose rate 192Ir source,” Physics in Medicine and Biology, vol. 46, no. 3, pp. N79–N90, 2001. View at Publisher · View at Google Scholar · View at Scopus
  15. P. Papagiannis, A. Angelopoulos, E. Pantelis et al., “Dosimetry comparison of 192Ir sources,” Medical Physics, vol. 29, no. 10, pp. 2239–2246, 2002. View at Publisher · View at Google Scholar · View at Scopus
  16. G. Douysset, T. Sander, J. Gouriou, and R. Nutbrown, “Comparison of air kerma standards of LNE-LNHB and NPL for 192Ir HDR brachytherapy sources: EUROMET project no 814,” Physics in Medicine and Biology, vol. 53, no. 6, pp. N85–N97, 2008. View at Publisher · View at Google Scholar · View at Scopus
  17. S. J. Goetsch, F. H. Attix, D. W. Pearson, and B. R. Thomadsen, “Calibration of 192Ir high-dose-rate afterloading systems,” Medical Physics, vol. 18, no. 3, pp. 462–467, 1991. View at Publisher · View at Google Scholar · View at Scopus
  18. ICRU, “Stopping powers for electrons and positrons,” ICRU Report 37, 1984.
  19. D. W. O. Rogers, I. Kawrakow, J. P. Seuntjens, et al., “NRC user codes for EGSnrc,” NRCC Report PIRS-702(revB), 2005.
  20. X-5 Monte Carlo team, MCNP-A General Monte Carlo N-Particle Transport Code, version 5, Los Alamos National Laboratory, Los Alamos, NM, USA, 2003.
  21. D. W. O. Rogers and J. Borg, “Spectra and air-kerma strength for encapsulated 192Ir sources,” Medical Physics, vol. 26, no. 11, pp. 2441–2444, 1999. View at Publisher · View at Google Scholar · View at Scopus
  22. S. M. Seltzer and P. M. Bergstrom, “Changes in the U.S. primary standards for the air kerma from gamma-ray beams,” Journal of Research of the National Institute of Standards and Technology, vol. 108, no. 5, pp. 359–381, 2003. View at Scopus