INTRODUCTION
In radiotherapy treatment planning, computed tomography (CT) images are used to calculate the dose distribution in the patient. In patients with a high electron density hip prosthesis, however, the prosthesis can cause streaking artefacts in the images, making dose calculations for nearby organs complicated and inaccurate. One solution to this problem is to avoid directing the photon beam towards the prosthesis, but this is not always feasible. When it is possible to avoid the hip prosthesis, doses reaching organs at risk, such as the rectum and the bladder, may be higher.Reference Palleri, Baruffaldi, Angelini, Ferri and Spezi 1
To better understand the photon interaction with high-density metal, an accurate radiation interaction model is needed. In this model, the real tissue composition should be estimated instead of using water with different densities to mimic all tissue types.Reference Reynaert, Van de Marck and Schaart 2 Conventional radiation interaction models fail to eliminate the uncertainties originating from streak artefacts in CT images.Reference Abdul Aziz, Yusoff, Osman, Abdullah, Rabaie and Salikin 3 , Reference Bazalova, Beaulieu, Palefsky and Verhaegena 4 Streak artefact reduction techniques have been discussed comprehensively in the literatureReference Abdul Aziz, Yusoff, Osman, Abdullah, Rabaie and Salikin 3 , Reference Osman, Salikin, Saripan, Aziz and Daud 5 , Reference Abdoli, Mohammad Reza, Ahmadian, Sahba and Zaidi 6 by using the dual step adaptive thresholding technique, the sinogram correction method, and filtering methods to remove the artefacts from affected CT images.
Experimental methods using a dosimeter are the most accurate techniques for quantifying the radiation dose in the medium and for understanding the fundamental interactions between photons and the medium. Phantoms composed of different materials are commonly used in place of human tissues in radiation studies, but the dimensions of phantoms and dosimeter factors can affect the data. In addition, limited data make it difficult to establish accurate radiation interaction models. In situations in which experimental protocols are not feasible, Monte Carlo (MC) simulation, which is known for its dose calculation accuracy, can be used to develop an effective model. MC is one type of computational algorithm that is based on repeated random sampling to compute the results. MC dose simulation has been reported to provide low uncertainty calculation (<2%) in extreme inhomogeneities dose calculation.Reference Abdul Aziz, Yusoff and Salikin 7 – Reference Toossi, Ghorbani, Akbari, Sabet and Mehrpouyan 9 In short, when the measurements of MC dose planning is carefully benchmarked, it can provide a valuable additional tool for treatment planning when difficult measurement situations occurr.Reference Reynaert, Van de Marck and Schaart 2
In the present study, dosimetric calculation was performed to assess the effects of various hip prosthesis materials on 6 MV dose distributions using a MC source code (BEAMnrc and DOSXYZnrc). MC simulation was able to do a calculation in various materials and geometries without requiring extensive experimental measurements to be taken. The results of this study serve as a guideline for dose calculation whenever the primary beam must be directed at the hip prosthesis materials.
MATERIALS and METHODS
Phantom scanning and image processing technique
A phantom containing Perspex and Teflon to represent soft tissue and the femoral head of the human hip, respectively, was prepared for use in this study (Figure 1a). The phantom (total size 14×28×22 cm) was scanned using a Toshiba CT simulator system (Toshiba Aquilion 64, Toshiba Medical Systems, Tokyo, Japan) to obtain the tomographic images (Figure 1b). The phantom was custom made based on Asian male pelvic size and consisted of Perspex (1·18 g/cm3) to represent human tissue and Teflon (2·20 g/cm3) to represent the head of the femur. To quantify the radiation dose at the hip prosthesis accurately, CT images should be free from the influences of metal streak artefacts. For this reason, image processing techniques were used to build computer-generated phantoms consisting of different prosthesis materials and densities substituting the Teflon femur head. CTCREATE software was used to convert the Teflon’s density to that of various hip prosthesis materials using the data listed in Table 1.
Figure 1 (a) A phantom made from Perspex and Teflon used in this study. (b) An image of the phantom scanned using a 64 multislice computed tomography simulator system. Notes: Teflon can be observed clearly in the image.
Table 1 Density and composition of materials used in the DOSXYZnrc simulation
MC modelling, simulation and beam setup
The MC technique was used in this study to quantify the dose distribution in the Perspex/Teflon hip prosthesis and in the computer-generated phantoms. The MC model used in this study was obtained from the International Atomic Energy Agency (IAEA) MC database for the linear accelerator (Siemens Primus, Siemens Medical Systems, Concord, CA, USA), which contains validated accelerator beam data (phase-space file) from various model types, beam energies and field sizes. The phase-space data were executed in BEAMnrc code, and the output of BEAMnrc source code was used to calculate the dose distribution in the DOSXYZnrc source code. This study was performed using a 6 MV photon beam, 10×10 cm2 field size and 100 cm source to surface distance (SSD). The beam was directed lateral to the phantom as shown in Figure 2 to determine the maximum perturbation that usually occurs in the clinical setting. The number of histories used in DOSXYZnrc was 150 million (~1 hour calculation time). MC transport parameters (photon cut-off energy and electron cut-off energy)were set to 0·01 and 0·7 MeV, respectively.
Figure 2 Photons with energy 6 MV were directed lateral to the phantom. Notes: The first and second Teflon pieces are located 4 and 19 cm, respectively, from the surface.
RESULTS and DISCUSSION
Siemens Primus model validation
The spectral distribution was plotted at the phantom surface with SSD 100 cm for the 10×10 cm2 field size (Figure 3). The average photon energy of the 6 MV photon beam was approximately one-third of the maximum nominal energy. The energy spectra of incident photon peaks were found at 0·5 MeV, which is similar to that previously reported by Aljamal and Zakaria.Reference Aljamal and Zakaria 8 The phase-space file downloaded from IAEA MC database was confirmed as 6 MV photon beam file due to the maximum photon energy observed as 6 MeV in the spectrum. Scatter plot analysis also shows that the field size was 10×10 cm2 for both the X and Y axis (Figure 4).
Figure 3 Spectrum of 6 MV photons from the International Atomic Energy Agency phase-space database (Siemens Primus).
Figure 4 Scatter plot analysis illustrates a precise 10×10 cm2 field size of the simulated model.
MC dose calculation
The DOSXYZnrc was used to quantify the dose distribution of the 6 MV photon beam from the Siemens Primus linear accelerator. As a reference, the plot of percentage depth dose (PDD) of water (Figure 5) shows that the dose reached its d max at 1·5 cm depth. After that, the dose decreased significantly until it reached the exit point of the phantom. PDD and beam data were used to analyse the impacts of the hip prosthesis on the dose distribution of the photon beam. PDD data showed that the absorbed dose decreased as the photon beam struck the higher density materials (bone, titanium, stainless steel and CoCrMo). The dose also increased minimally at both interfaces of Perspex high-density regions.
Figure 5 Percentage depth dose of the 6 MV photon beam oriented transverse to the various materials with different densities [water, bone, titanium (Ti), stainless steel (SST) and CoCrMo]. Notes: Drastic perturbation of doses were observed at depths of 4–9 cm (first femoral head position) and 19–24 cm (second femoral head position). Water ρ, 1 g/cm3; bone ρ, 2·08 g/cm3; Ti ρ, 4·48 g/cm3; SST ρ, 6·45 g/cm3; and CoCrMo ρ, 8·20 g/cm3.
Further analysis showed that as the depth increased, the absorbed doses decreased by different magnitudes depending on the density of the material and the location. The dose was least perturbed at 0·5 cm depth and most perturbed at depth 9·5 cm from the surface. Thus, when a high-density prosthesis is located between beam and tumour, the beam weighting must be increased to deliver a sufficient dose to the tumour (i.e., the prostate) but at the same time the hot spot at the surface must be avoided. Of the high-density materials tested, the CoCrMo hip prosthesis material had the highest density, and it perturbed more radiation compared with bone, titanium and stainless steel. Figure 6 shows that the magnitude of dose perturbation was directly proportional to the density of the material struck by photons. Bone with density of 2·08 g/cm3 was the least perturbed material followed by titanium (4·48 g/cm3) and stainless steel (6·45 g/cm3), with 32 and 41% dose difference relative to water, respectively. For CoCrMo, the dose difference was 49%. Overall, the contribution of backscatter to the PDD curve was minimal at the interfaces of the phantom (i.e., surrounding tissue).
Figure 6 Percentage difference of dose perturbation of the 6 MV photon beam oriented transverse to various materials with different densities [water, bone, titanium (Ti), stainless steel (SST) and CoCrMo]. Notes: Bone, titanium and stainless steel effects on dose distribution were as high as 7, 32 and 41%, respectively, compared with water. The highest perturbation was observed for CoCrMo, for which a 49% difference was calculated.
CONCLUSIONS
The presence of a high-density hip prosthesis in a patient makes it difficult to calculate the dose distribution prescribed for the target and the surrounding tissues in treatment planning. The PDD and beam profile generated in this study for several types of hip prosthesis materials showed that high-density materials attenuated the photon beam to a significant extent relative to water. With an increase in the density of the material transversed by the 6 MV photon beam, attenuation of the beam increased and perturbation of the dose increased, resulting in a dose reduction as high as 49% for the densest (CoCrMo) material tested. The technique used in this study successfully quantified the perturbation effect on a dedicated pelvic phantom, whereas excluding the potential dose uncertainties arising from streak artefacts on CT images.
Acknowledgements
The authors would like to extend their gratitude to the Department of Higher Education (Malaysia) for providing a grant under Fundamental Research Grants Scheme (FRGS) 203/CIPPT/6711341. Appreciation also goes to the staff of the Radiotherapy Unit of the Advanced Medical and Dental Institute at Universiti Sains (USM) the School of Physics at USM, Radiotherapy and Oncology Department at Universiti Sains Malaysia (USM) and the staff of the Radiology Department at Universiti Sains Malaysia (USM).
