Tissue equivalence of soe phanto aterials for proton beas V.N. Vasiliev 1*, V.I. Kostjuchenko 2, O.B. iazantsev 2,V.G. Khaybullin 2, S.I. Saarin 3, A.S. Uglov 3 1 Institute for Nuclear esearch, ussian Acadey of Sciences, Mosco, ussia 2 Institute for Theoretical and Experiental Physics, Mosco, ussia 3 ussian Federal Nuclear Center Zababakhin Institute of Applied Physics, Snezhinsk, ussia Abstract: Tissue and ater equivalence of soe phanto aterials originally developed for conventional radiation therapy as investigated on the ITEP edical proton bea facility. The proton CSDA range in three variants of Plastic Water, lung, adipose, uscle and copact bone substitute aterials (CIS Inc., USA) as ured by a silicon diode as ell as the residual proton range in liquid ater after passing a slab of each aterial under investigation. In addition, the proton range in five aterials of knon eleental coposition as calculated by Monte Carlo technique. The obtained results ere copared ith erence data fro ICU report 49 for respective biological tissues and ater. A total uncertainty of the proton range ratios as estiated to be fro 0.9 to 1.5% (1SD). Within these uncertainties, Plastic Water, Plastic Water L, Plastic Water DT, uscle and copact bone deonstrated a good agreeent ith the erence data. The range in adipose and lung substitutes is a fe percents loer than that in the respective tissues. Introduction Application of proton therapy in edical practice requires a special equipent to siulate a huan body, organs and tissues as ell as the erence ediu, ater. This equipent includes soe dosietry phantos and test objects and has to be used both in bea coissioning and dose distributions verification process. The proton interaction cross sections of the phanto aterials are to be close as possible to those of respective biological tissues. This proble is ell knon in conventional radiation therapy by photon and electron beas. Nuerous tissue and ater equivalent aterials ere developed for radiation therapy last decades [1-3] and high accuracy in the interaction cross sections siulation as achieved for odern substitutes. In particular, ater and tissue equivalent plastics anufactured by CIS Inc. are close to liquid ater and respective tissues ithin 0.5-1.0% for photon and electron beas. Nevertheless, the applicability of these aterials in proton beas is to be specially verified. In this ork, e have perfored an experiental and theoretical evaluation of tissue and ater equivalence of seven phanto aterials originally developed for conventional radiation therapy. A proton range ratio ured in each aterial under investigation and in liquid ater as used as an equivalence estiator and copared ith theoretical values for respective tissues provided by ICU eport 49. In addition, proton ranges in four plastic saples of knon eleental coposition ere siulated by Monte Carlo technique. The results of the coparison alloed evaluating the equivalence of these plastics respective to proton range and stopping poer. Other significant paraeters, the scattering poer and the inelastic nuclear reaction contribution, are planned for investigation in the future. Materials and ethods Three types of Plastic Water (developed for high, ediu and lo photon energy), adipose, uscle, lung and cortical bone substitutes (CIS Inc., USA [4]) ere used in our ureents for proton range estiation. All saples ere anufactured as 10x10 c slabs of 1, 5, 10 and 20 * E-ail: vnvasil@orc.ru 1
thickness, an additional 20 slab as intended as an adapter plate and had a cylindrical cavity for a detector. A difference beteen noinal and real thickness of the slab stack as less than 0.3% except for 0.6% for the uscle saple. The ureents ere perfored on a horizontal edical proton bea of the ITEP synchrotron facility. The proton bea as spread out by the double scattering technique ith a profiled secondary scatterer for better lateral fluence unifority (see, for exaple, [5]) and passed through ater bellos used as an energy degrader. The bea diaeter as liited by a 7 c steel colliator placed before the energy degrader. A ogovsky coil as used as a bea onitor, all dose values ured by the detector ere noralized to the coil reading to account for the bea instability. All dose ureents in plastics and liquid ater ere perfored ith a silicon diode. Depending on the aterial under investigation, three experiental setups ere used as shon in Fig. 1. Setups A and B ere used for the direct CSDA range ureent and subsequent coparison in plastic saples and ater. An initial energy of the proton bea as 219 MeV, and a ater thickness in the energy degrader as 100. For the plastic ureents (Setup A), the detector as placed in a cavity inside the adapter plate at a distance of 170 fro the energy degrader indo. The investigated plastic slabs have been added in front of the adapter plate by steps fro 20 at the dose plateau to 1 at the Bragg peak. As the necessary thickness of the lung saple is too high to achieve the Bragg peak due to its lo physical density (up to one eter!), it as ured in cobination ith 13 c of Plastic Water DT. In that case, the energy degrader to detector distance as 30 c to place the slabs of both types. The proton range in liquid ater (Setup B) as ured in a ater filled PMMA phanto by a 3D scanning device for detector positioning (Fig. 1b). The phanto as set closely to the energy degrader output indo; the thickness of the phanto entrance indo as 2.5 of PMMA and equivalent to 2.9 of ater, taking into account their linear stopping poer ratio. The data as corrected for that thickness in further processing. As the geoetries of ureent in ater and plastic (Setup A and B) are different, soe correction for a bea divergence and proton fluence change as applied to in-ater results. Nevertheless, this correction had only negligible influence, less than 0.1, on the estiated proton range. Setup C (Fig.1c) as used for estiation of the residual proton range in ater after passing 160 of the investigated plastic (120 for cortical bone). Plastic slabs of that total thickness ere placed beteen the energy degrader output indo and the ater filled phanto. For coparison, one ureent as perfored in ater ithout the plastic, setting the phanto close to the energy degrader. The initial bea energy in these ureents as 220 MeV, the ater thickness in the energy degrader as reduced to 20 to ensure enough residual range in ater. Finally, depth dose distributions in three types of Plastic Water and cortical bone ere calculated by Monte Carlo technique. The proton transport siulation as perfored by the progra IThMC developed for proton therapy planning and alloing a dose calculation in voxel geoetry (up to 512x512x512 voxels). The progra takes into account the ionization energy loss in the ediu, the energy straggling (by the Landau, Vavilov or noral distributions depending on the aterial thickness), the elastic ultiple coulob scattering using the Fokker- 2
Planck and Feri-Eyges odels, the elastic and inelastic nuclear reactions on the base of the Sychev odel and the D2N2 cross section data set respectively [6]. The siulation geoetry as siilar to Setup A/B but soehat siplified. A parallel bea of 200 MeV protons ith E/E = 0.6% passed through a 100 ater slab siulating the ater energy degrader and a 300 slab of the aterial under investigation. Transport of 10 7 incident protons as siulated for each aterial. Depth dose data ere calculated along the bea axis using a voxel size of 1 and then CSDA ranges ere derived fro obtained results. Fig. 1. The ureent setup: a direct range ureents in the plastic slabs; b direct range ureents in liquid ater; c residual range ureents in ater after the plastics passing. 1 the proton bea transport syste; 2 the priary scatterer; 3 the steel colliator; 4 the secondary profiled scatterer; 5 the ater bellos (energy degrader); 6 the saple slabs; 7 the adapter plate; 8 the diode; 9 the ater filled phanto; 10 the 3D detector positioner; esults Direct ranges coparison Measured depth dose distributions in all plastics using Setup A and in ater using Setup B are shon in Fig. 2. The corrections for proton bea divergence and the phanto all ere applied 3
to the ater data as described above. The CSDA range as estiated as the depth distal to the Bragg peak here the dose is reduced to 80% of its axiu value. Mean proton energy at the surface of investigated saples calculated on the base of ICU49 [3] range-energy relation as estiated to be 153.9 MeV. All data are noralized to the Bragg peak axiu. Soe difference at the dose plateau can be resulted fro an uncopensated contribution of the secondary particles generated in the inelastic nuclear reactions. 1.2 1.0 Dose, arb.units 0.8 0.6 0.4 0.2 0.0 0 5 10 15 20 Depth, c a) PWL PWDT PW CIS Adipose CIS Muscule Water CIS Cortical bone 1.2 1.0 Dose, arb.units 0.8 0.6 0.4 0.2 b) 0.0 13 14 15 16 17 18 Depth, c PWL PWDT PW CIS Adipose CIS Muscule Water Fig. 2. Measured depth dose distributions in plastics and ater: a the full curves, b the Bragg peaks. 4
To estiate the ater and tissue equivalence of the investigated aterials, the ured proton ranges ere copared ith respective erence data taken ainly fro ICU eport 49 [3]. In addition, as lung data absent in eport 49, eleental copositions of lung and cortical bone ere obtained fro the Woodard and White paper [7] and used for proton range calculations ith the progra SIM [8, 9]. There is a little discrepancy beteen these to erence data sets resulted fro slightly different excitation potential and the shell correction as discussed in [10, 11]. This discrepancy, as a rule, is ithin their estiated uncertainties and, for exaple, is about 1.4% for proton range in ater at 150 MeV. Nevertheless, to avoid the influence of this disagreeent, all ranges to copare ere related to those in liquid ater, experiental or theoretical. As the result, folloing estiator of the tissue/ater equivalence as used: / C =, (1) / tissue here and are ured CSDA proton ranges in the tested substitute and in ater; tissue is erence proton range tabulated for the respective tissue (ICU49 or SIM); erence proton range in ater. The results of the first ureent series are presented in Table 1. Table 1. Proton range coparison fro direct ureents. Substitute ρ, g/c 3, c / C Data set The aterial for coparison is Plastic Water 1.030 16.22 1.013 1.013 ICU49 Water Plastic Water 1.029 16.19 1.010 1.010 ICU49 Water L Plastic Water 1.039 16.47 1.028 1.028 ICU49 Water DT Cortical bone 1.91 9.72 0.606 1.004 ICU49 Cortical bone ICP, ρ =1.85 g/c 3 Cortical bone 1.91 9.72 0.606 1.042 ICU49 Copact bone ICU, ρ =1.85 g/c 3 Cortical bone 1.91 9.72 0.592 1.024 SIM Woodard-White bone, ρ =1.93 g/c 3 Adipose 0.96 16.60 1.036 0.981 ICU49 Adipose ICP, ρ =0.92 g/c 3 Muscle 1.06 15.58 0.972 1.000 ICU49 ICP Skeletal Muscle, ρ =1.04 g/c 3 Muscle 1.06 15.58 0.972 1.002 ICU49 ICU Striated Muscle, ρ =1.04 g/c 3 Lung +13 c of PW DT 0.205 16.48 4.914 0.974 SIM Woodard-White lung, ρ =0.200 g/c 3 5
esidual range coparison The second series of ureents as perfored by Setup C (Fig.1c) and alloed to estiate the residual proton range in ater after passing 120 (for cortical bone) or 160 (for other saples) of the plastic under investigation. Subtracting the obtained residual range fro that in ater ithout plastic, an equivalent ater thickness as estiated, related to the plastic thickness and then copared ith erence range ratio. Thus, in this series of ureents, the tissue/ater equivalence estiator C as defined as C T /( + ) =, (2) tissue / here T is the thickness of plastic slab, and + are the ured proton range in ater ithout and ith plastic slab, and tissue are the erence range in ater and in the respective tissue. An advantage of this ethod is the identity of range ureent conditions for plastic and liquid ater. The obtained results are presented in Table 2. Table 2. esidual ranges coparison. Substitute T, +, T C Data set The aterial for c c ( + ) coparison Plastic Water 15.94 7.58 0.987 0.987 ICU49 Water Plastic Water 15.98 7.60 0.991 0.991 ICU49 Water L Plastic Water 16.03 7.66 0.997 0.997 ICU49 Water DT Cortical bone 11.93 3.54 0.591 0.981 ICU49 Cortical bone ICP, ρ =1.85 g/c 3 Cortical bone 11.93 3.54 0.591 1.017 ICU49 Copact bone ICU, ρ =1.85 g/c 3 Cortical bone 11.93 3.54 0.591 1.000 SIM Woodard-White bone, ρ =1.93 g/c 3 Adipose 15.97 8.10 1.021 0.966 ICU49 Adipose ICP, ρ =0.92 g/c 3 Muscle 15.89 7.00 0.950 0.978 ICU49 ICP Skeletal Muscle, ρ =1.04 g/c 3 Muscle 15.89 7.00 0.950 0.980 ICU49 ICU Striated Muscle, ρ =1.04 g/c 3 Lung 15.95 20.36 4.738 0.941 SIM Woodard-White lung, ρ =0.200 g/c 3 Monte Carlo siulation As described above, an axial depth dose distribution as calculated in four aterials by Monte Carlo technique using the IThMC proton transport code. The geoetry of siulation as siilar to Setup A/B (Fig.1a and 1b); theore, the estiator C according to equation (1) as used for the ater/tissue equivalence verification. The siulation results are shon in Table 3. 6
Table 3. Monte Carlo siulation results. Substitute ρ, g/c 3, c / C Data set The aterial for coparison Plastic Water 1.03 15.91 1.003 1.003 ICU49 Water Plastic Water 1.029 15.89 1.002 1.002 ICU49 Water L Plastic Water 1.039 15.89 1.002 1.002 ICU49 Water DT Cortical bone 1.91 9.49 0.598 0.991 ICU49 Cortical bone ICP, ρ =1.85 g/c 3 Cortical bone 1.91 9.49 0.598 1.028 ICU49 Copact bone ICU, ρ =1.85 g/c 3 Cortical bone 1.91 9.49 0.598 1.010 SIM Woodard-White bone, ρ =1.93 g/c 3 Error estiation The proton range reproducibility as estiated by repeated ureents in ater under the sae bea paraeters. A range standard deviation as 0.7, it as introduced by the detector positioning accuracy. The plastic slabs thickness and respective detector position in plastic as ured essentially ore accurate, the thickness uncertainty never exceeded 0.02. Another significant contribution to the range uncertainty arose fro the proton energy instability. A typical energy scatter as 0.5 MeV (i.e. about 0.2%) and resulted in a 1.2 proton range error (1SD). Statistical fluctuations of the bea onitor and the diode response resulted in about 0.1 of the proton range uncertainty (1SD) calculated by the error propagation forula. Thus, a total uncertainty of the proton range in ater and plastic as estiated to be 1.4 and 1.2 (1SD) respectively. That leaded to an uncertainty of the plastic to ater proton range ratio fro 1.1 to 1.5% depending on the plastic type in the first ureent series (the direct range coparison) and fro 0.9 to 1.3 % in the second ureent series (the residual range coparison). The only exception as the lung equivalent saple in the second series, here the uncertainty reached 5.6%. Conclusions A coparison of Table 1 and 2 deonstrates soe systeatic difference beteen the results obtained in direct and residual ranges ureents. Average discrepancy is about 2.5% and can be explained by the ureent uncertainty as estiated above, at least, ithin ±2SD confidence liits. A systeatic nature of this difference allos suggesting the influence of the proton energy instability. A contribution of that factor is to be iniized in future investigations to iprove the total accuracy of the results. The proton ranges in all Plastic Water aterials are close to those in liquid ater. Theoretical values ratio calculated by Monte Carlo technique is unity ithin 0.3%. A coparison of the residual ranges in ater deonstrates a 0.3-1.2% difference against the erence value. Direct range coparison shos orse agreeent 1.0-1.3% and up to 2.8% for PW DT. The last is the only result lying out of the ±2SD confidence interval. 7
The range in uscle substitute as copared ith that in ICP skeletal uscle and in ICU striated uscle. The deonstrated differences ere fro 0.0-0.2% to 2.0-2.2% in second and first ureent series respectively. Soe proton range underestiation as obtained in the adipose substitute fro 1.9% in the first series to 3.3% in the second one. The last value is situated out of the ±2SD interval. Confority of the cortical bone results essentially depends on the erence data used for coparison. ICU copact bone (ICU49), ICP cortical bone (ICU49) and Woodard-White bone [7] sho a 3.6% value scatter. As the result, the proton range discrepancy as 0.4-4.2% and -0.9-+1.7% in the first and second ureent series respectively. A proton range in lung as underestiated against the Woodard-White/SIM data in both series 2.6% in the first and 5.9% in the second one. Nevertheless, the conditions of the second ureent series deonstrated a bad sensitivity in lo density aterials, like lung, and, respectively, the estiated standard deviation aounting to 5.6%. Thus, ithin the estiated uncertainties, Plastic Water, Plastic Water L, Plastic Water DT, uscle and copact bone deonstrated good agreeent ith liquid ater and respective tissues relative to proton range and stopping poer. The range in adipose and lung substitutes is a fe percents loer than that in the tissues. eferences 1. White D.., Constantinou C. Anthropoorphic phanto aterials. In: Progress in Medical adiation Physics, 1982, vol. 1, 133-193. 2. Tissue substitutes in radiation dosietry and ureent. ICU eport 44, Bethesda, 1989. 3. Stopping poers and ranges for protons and alpha particles. ICU eport 49, Bethesda, 1993. 4. Plastic Water. CIS Catalog. http://.cirsinc.co. 5. Gottschalk B. Passive bea spreading in proton radiation therapy. Harvard High Energy Physics Laboratory, 2004. http://huhepl.harvard.edu/ gottschalk. 6. Sychev B.S. High energy hadrons interaction cross sections ith atoic nuclei. Mosco adiotechnical Institute, Mosco, 1999 (in ussian). 7. Woodard, H.Q., White, D.. The coposition of body tissues. Brit. J. adiol. (1986) 59: 1209-1219. 8. Ziegler J.F., Ziegler M.D., Biersack J.P. SIM. The stopping and range of ions in atter. http://.sim.org. 9. Ziegler J.F. The stopping of energetic light ions in eleental atter. J.Appl.Phys. (1999) 1249-1272. 10. Ziegler J.F. Coentary: Coents on ICU eport 49: Stopping poers and ranges for protons and alpha particles. adiat.es. (1999) 152, 219-222. 11. Seltzer S.M., Inokuti M., Paul H., Bichsel H. esponse to the coentary by J.F.Ziegler regarding ICU eport 49: Stopping poers and ranges for protons and alpha particles. ICU Nes (2001) No.2, 1-10. 8