comparison of various methods for designing the

Radiation Protection Dosimetry Advance Access published August 16, 2012 Radiation Protection Dosimetry (2012), pp. 1–5

doi:10.1093/rpd/ncs173

COMPARISON OF VARIOUS METHODS FOR DESIGNING THE SHIELDING FROM IONISING RADIATION AT PET-CT INSTALLATIONS Vojislav Antic´1, Koviljka Stankovic´2, Milosˇ Vujisic´2 and Predrag Osmokrovic´2,* 1 Clinical Center of Serbia – Nuclear Medicine Center, Visegradska 26, Belgrade, Serbia 2 Faculty of Electrical Engineering, University of Belgrade, Bul. kralja Aleksandara 73, Belgrade, Serbia

Received April 22 2012, revised July 17 2012, accepted July 19 2012 Protection at positron emission tomography-computed tomography (PET-CT) installations is the most complex problem in the field of designing structural protection from ionising radiation in medical practice. This paper provides a discussion on the values for shield widths obtained from two different estimation methods, as well as of certain theoretical differences inherent in the two approaches. After the general operation principles of a PET-CT device are expounded, a comparative analysis of two methods for calculating structural barriers is performed. The first calculation was conducted by the ‘Vincˇa’ Institute of Nuclear Sciences, according to the recommendations of the AAPM task group 108, while the second was performed by a PET-CT device manufacturer, following the DIN 6844-3 standard.

INTRODUCTION A typical design of a positron emission tomographycomputed tomography (PET-CT) device combines a CT scanner gantry and a ring-shaped PET detector system within the same housing, with a single patient table. CT provides high-quality anatomical images of transverse cross-sections of the body, while PET gives functional information—metabolic activity of certain tissues or their parts, in relatively low resolution, as far as localisation of lesions is concerned, and without any tissue-specific amplification correction. By fusing these two imaging techniques, a unique medical image is obtained, which contains both morphological and functional information on the investigated tissue. Basic principles of PET-CT image generation are shown in Figure 1. Radionuclides used in PET procedures are shortlived bþ emitters, produced in cyclotrons. The majority of procedures use radioactive fluorine 18F obtained 18 from the nuclear reaction 18 8 O þ p ! 9 F þ n, and 18 0 þ F ! O þ b þ n with a half-life of decays as 18 9 8 1 T1=2 ð18 FÞ 110 min. An appropriate radiopharmaceutical (in case of 18F it is the FDG, i.e. 2-deoxy-2(18F)fluoro-D-glucose) is then produced in a synthesis module. Upon intravenous injection, a high concentration of the radiopharmaceutical accumulates in metabolically active tumours (accumulation time for FDG is 1 h). At the physical level, while a positron transverses the body (with a range of 1 mm), its energy decreases due to Coulomb interactions, it then ‘captures’ an electron, which results in a positron – electron annihilation, producing two ‘annihilation’ gamma photons emitted in opposite directions.

Simultaneous (coincidental) detection of such a photon pair in corresponding opposite detectors represents basic diagnostic information. The signal is amplified in photomultipliers, led through optical cables and fed to an electronic/computer system which displays the place (voxel) of the decay and the activity in it. CT uses an X-ray tube that rotates constantly and emits X rays, while the patient is constantly moved through the centre of the gantry. Different contrasts on the image are obtained on the basis of CT numbers, ascribed to various tissues according to attenuation of X rays they produce. As was previously said, PET-CT provides both functional and anatomical diagnostic information, but since it uses ionising radiation it is necessary to mitigate the inevitable exposure to radiation and completely reduce unnecessary exposure. The three main principles of ionising radiation protection in radiation practices are the justification of practice, optimisation of protection and limitation of individual exposure(2). While classic nuclear medicine mainly utilises the radionuclide 99m Tc, which emits 140 keV photons by relaxing from a metastable to the ground state, PET annihilation photons have an energy of 511 keV. As for the CT component, maximum effective mean energy of X rays does not go over 70 keV (in a CT, the so-called secondary radiation is a sum of the radiation leaking from the X-ray tube housing and the scattered radiation arising from photon interactions with the patient’s tissue, the detector system and other objects in the primary beam). A diagram of the scattered radiation (isodose curves) is provided

# The Author 2012. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

Downloaded from http://rpd.oxfordjournals.org/ at University Library Svetozar Markovic on October 3, 2012

*Corresponding author. [email protected]

´ ET AL. V. ANTIC

DETERMINATION OF TRANSMISSION COEFFICIENTS IN PET Transmission coefficient describes the amplitude, intensity or total power of the transmitted wave with respect to the incident wave. In dosimetry and radiation protection, the transmission factor is defined as Figure 1. The mechanism of PET-CT image generation(1).

B¼

ð1Þ

where DðdÞ is the dose in the observed point with the protection present, whereas D0 ðdÞ is the dose in that same point without protection. In view of the estimations and calculations in ref.(2), the transmission factor is obtained from the following formula: B¼

Pd 2 GTNA0 tu Rt

ð2Þ

where G is the effective dose equivalent. AAPM task group 108 proposes 0.092 mSvm2 MBqh21 as its value for the calculation. According to ANS-1991, its initial value is taken to be 0.143 mSvm2 MBqh21. This value has to be corrected for attenuation in the patient’s body, which is assessed by the AAPM task group 108(3) to be 0.36 (effective absorption in the body), on the basis of many studies for different patient orientations(8 – 13). According to the calculation in ref. (14), the total body absorption factor at 500 keV is 0.34, which coincides well with the assessed value. Hence, by simply multiplying the initial value by 0.64, the proposed value of the dose equivalent is obtained. P is the weekly dose limit. It is 20 mSv, considering the annual limit for the members of the public (1 mSv), or 100 mSv considering the annual ‘target’ level in compliance with local legislation (5 mSv according to International Atomic Energy Agency (IAEA) recommendations) for radiation workers. d is the distance from the source to the position behind the barrier where measurement is performed. T is the retention factor. It has a value between 0 and 1, depending on the assessment of effective irradiation of the most exposed individual in the area or at a point of interest, or it can be set in compliance with legislation. N is an assessed number of patients undertaking the examination per week. A0 is the ‘administered’ activity, designated to be applied. Task group 108 adopts the value of 555 MBq 18F FDG for calculations(3). t is the time interval which is of interest for a specific calculation (application time, survey time). Rt is the dose reduction factor. Because PET tracers have short half-lives, the total radiation dose received over a time period t, D(t), is less than the product of the initial dose rate and time D0 ð0Þt.

Page 2 of 5


by the manufacturer, and it represents a starting point for dose assessment and design of appropriate protection in CT diagnostics. Although there are multiple contributions of CT to the study of a PETCT system (formation of a topogram, attenuation correction and image fusion), X-ray energies of a CT in a PET-CT configuration are considerably lower than in classic CT diagnostics. American Association of Physicists in Medicine (AAPM) has published a paper in 2005 on the structural protection for PET-CT installations, titled ‘PET/CT Shielding Requirements’(3). The postulates in this study are adopted by most practical solutions, as a basis for calculating the properties of structural protection. There are certain deviations, both in theory and in practical calculations. One specifically considered example, regarding clarifications of the problem and pointing out the differences, is the project done for the National PET Center at the Clinical Center of Serbia. The device whose installation is planned is a Siemens PET-CT model Biograph True point 40/64(4). In accordance with a defined working procedure, the user has designated a room plan for the PET-CT installation. Calculation of barriers according to the AAPM task group 108(3) has been made by the ‘Vincˇa’ Institute of Nuclear Sciences(5). On the other hand, the manufacturer has produced a worksheet based on the DIN 6844-3 standard(6, 7), containing another calculation of ionising radiation protection for this device. Protection at PET-CT installations is the most complex problem in the field of designing structural protection from ionising radiation in medical practice. This paper, therefore, provides a comparison of different methods for estimating structural protection. Principles of both calculation methods (one made by the ‘Vincˇa’ Institute of Nuclear Sciences according to the recommendations of the AAPM task group 108, and the other by the manufacturer based on the DIN 6844-3 standard) are considered in the paper, with direct implementation to the premises of the National PET Center of Serbia. Medical practice within the area that provides parameters necessary for the calculation is reviewed, along with certain theoretical differences, with the aim of contributing to the improvement of standards regarding optimisation of protection at PET-CT installations.

DðdÞ D0 ðdÞ

COMPARISON OF VARIOUS METHODS FOR DESIGNING

The reduction factor, Rt, is calculated as(3) Rt ¼

DðtÞ 1 elt ¼ D0 ð0Þt lt

ð3Þ

Calculation of protection has to be performed for all zones within and surrounding the PET installation. The distance behind a wall at which the value of the transmission factor is calculated is taken to be 0.3 m(3, 15 – 17). Dose rate for rooms above the installation is determined at a height of 0.5 m from the floor, and for rooms beneath the installation at a height of 1.7 m from the floor(3, 15 – 17). TENTH VALUE LAYER VS. MONTE CARLO On the basis of the B factor calculated from equation (2), the required width of the protective barrier is assessed. There are two fundamentally different methods. Assessment of structural protection according to the corresponding DIN standard 6844(6), or under NCRP (National Council of Radiation Protection) conditions(3, 15, 17) is based on the principle of attenuating the beam to one-tenth of its initial value using the so-called tenth value layers (TVLs), obtained through measurements with wide beams. TVL is the width of the material used for the barrier for which the radiation level becomes 10 times lower (i.e. reflected in dose reduction). For the monoenergetic beam with the energy of 511 keV, as well as for polyenergetic beam with effective energy of 511 keV, TVL for lead is 16.6 mm and for 2.35 g cm23 density concrete it is 17.6 cm. According to the Monte Carlo method, applied to a wide beam, which makes the basis for AAPM task group 108 calculations, for the typically used materials the desired width x is obtained from: g 1 B þ b=a ln ð5Þ x¼ ag 1 þ b=a where a, b and g are parameters obtained by the numerical Monte Carlo simulation(18 – 24). For example, for a wide beam of 511 keV photons in lead, these coefficients are (3,18): a ¼1.543 cm21, b ¼ 20.4408

DISCUSSION According to all principles stated in ref. (3), which have been laid out in the previous two segments of this paper, the corresponding calculation by the ‘Vincˇa’ Institute of Nuclear Sciences was produced(5). The weekly dose limit of equation (2) was applied conservatively (20 mSv, considering the 1 mSv limit for the members of the public). The retention factor of equation (2) was mainly set at the maximum value of 1. The time between the application and the survey (t in equation (2)) was taken to be 60 min (according to a procedure in the National PET Center, this time is 85 min). The survey time (t in equation (2)) was taken to be 30 min (in practice, its mean value is between 21 and 24 min (7 –8 beds), and maximum 48 min (16 beds), not including repeated surveys, which are rare and targeted). When the transmission coefficient is calculated, the time between the application and the survey has to be added to this time, because the decays keep occurring from the initial moment (‘administered’ activity A0 in equation (2)). The only significant difference from ref. (3) is the value of the ‘administered’ activity, taken by the ‘Vincˇa’ Institute of Nuclear Sciences to be 370 MBq, instead of 555 MBq as AAPM task group 108 recommended, which proved to be absolutely justified, since the optimised mean value in the practice of the National PET Center of Serbia is below 300 MBq. Results obtained according to Lerner(7), which are in compliance with DIN standard 6844(6), deviate from those of the ‘Vincˇa’ Institute of Nuclear Sciences, even if the same value of administered activity was applied (370 MBq). First of all, as discussed earlier, after the value of the transmission coefficient is estimated, different principles are used for determining barrier widths(3, 6). The dose limit is taken to be that for radiation workers, which is 6 mSv, according to Ciraj and Kosˇutic´(5). Recommendation of the AAPM task group 108(3) is 5 mSv, while the

Page 3 of 5


where l is the decay constant (in this case 0.693/110 min), whereas t is time interval between administration and undertaken examination. Walls are usually made of a combination of lead and concrete, but may contain other materials. Total transmission factor is found as a product of individual transmission factors, based on the widths of different materials comprising the barrier: Y Bi ð4Þ Btot ¼

cm21, g ¼2.136, whereas in concrete their values are a ¼0.1539 cm21, b ¼ 20.1161 cm21, g ¼2.0752. Comparing the plots representing the dependence of the transmission coefficient on the barrier width, obtained from the TVL principle and by the Monte Carlo method, it can be noted that, for lead, they are nearly identical up to the width of 10 mm(3, 6). As the width of the lead barrier is increased, the required width obtained from the TVL principle becomes larger than the value determined from the Monte Carlo method. In the case of concrete, there are also significant differences between the plots, with the width required for 511 keV photons again being larger according to the TVL criterion(3, 6). It is worth pointing out that the estimations for concrete were obtained for the density of 2.35 g cm23, which has to be borne in mind in the construction phase.

´ ET AL. V. ANTIC

(1) The administered activity is reduced by the fraction that the patient excretes, manly by urinating, which is typically 0.15–0.2 (e.g. for a ¼0.15, A0 ¼555 MBq, initial activity decreases for more than 83 MBq), i.e. the activity is calculated as AðtÞ ¼ A0 elt aA0

ð6Þ

(2) The value of G used in calculations is 0.139 mSvm2 MBqh21 (the effective dose rate constant, ICRP 74), which is close to 0.143 mSvm2 MBqh21 (recommendation of the AAPM task group, effective dose equivalent, ANS-1991). However, this study did not include attenuation in the patient’s body, which gives the value of 0.092 mSvm2 MBqh21(3). Variations in the G coefficient are often the only point of dispute in calculations. For instance, according to(24) the tissue dose constant should be used (0.148 mSvm2 MBqh21), while Martin(25) suggests the maximum dose (0.188 mSvm2 MBqh21, ANS-1977). According to the Law on Protection Against Ionizing Radiation and Nuclear Safety(2), other previously

used variants are possible, ranging from 0.134 mSvm2 MBqh21 (air kerma rate constant) to 0.183 mSvm2 MBqh21 (deep dose equivalent, ANS-1977). CONCLUSION In PET-CT installations, the result of a medical survey is a hybrid image, fused from positron emission tomography (PET) and computed tomography (CT). PET can discover regions of unusual metabolic activity, while CT is used for their anatomical localisation, as well as for attenuation correction of positron radiation. The main radiation component in a PET device are the 511 keV annihilation gamma photons, while X rays from the CT device have energies up to 140 keV. In this paper, the two calculations of barriers used for protection from ionising radiation were compared, one that follows recommendations of the AAPM task group 108, and the other compliant with the DIN 6844-3 standard. Both calculations show that the barriers have to be 10 times thicker for protection from positron emitters than from CT X rays. A detailed description of how the transmission factor is calculated for a PET device is included in the paper. Variations of all relevant physical, physicochemical and metabolic parameters are considered, as well as those characterising medical practice. After the transmission factors are determined, barrier widths are estimated, either by using the Monte Carlo method (according to the recommendations of the AAPM task group 108) or the ‘tenth value layers’ principle (according to the DIN 6844 standard). The design which provides radiation safety has been developed by the ‘Vincˇa’ Institute of Nuclear Sciences, completely in accord with all recommendations of the AAPM task group 108, with a conservative value of the dose limit (1 mSv). The calculation performed according to the DIN 6844 standard is not consistent with IAEA recommendations and finds little application in recent practice. It should be noted that the dose limit in the latter calculation (6 mSv) is not compliant with the national legislative limit (5 mSv) and the corresponding principles of radiological protection implementation in Serbia, which often adopt limits set for the members of the public when structural protection from ionising radiation is designed. Both standards, and the corresponding calculations, neglect the fraction the patient excretes before the survey (typically 15 –20 %), which should be subtracted from the administered activity when barriers (walls) of the rooms used for surveys are designed. ACKNOWLEDGEMENTS The authors would like to thank the National PET Center of Serbia for providing us with data and materials used for this research. The Ministry of

Page 4 of 5


calculation of the ‘Vincˇa’ Institute of Nuclear Sciences(5) adopted the limit for the members of the public (1 mSv). Dose reduction factor Rt has not been considered or included in the calculation. Projects of radiological protection(5, 7) also contain calculations of protective barriers for CT X rays. In both cases, comparison of project demands leads to a conclusion that the necessary barrier width is 10 times larger for protection against positron emitters. It should be noted that the energies of X rays are considerably lower in CTs that are incorporated in PET-CT studies than in diagnostic CTs, so that no additional protection is needed. Examples of differences in the calculations for PET-CT facilities are the following: calculations by the ‘Vincˇa’ Institute of Nuclear Sciences uses 19 mm of lead and 22 cm of concrete for protecting a critical wall of the exit waiting room, while calculations in ref. (7) uses only 10 mm of lead and 17.1 cm of concrete. An even more drastic difference is found in the control room: 19 mm of lead and 22 cm of concrete according to Ciraj and Kosˇutic´(5), whereas that of Lerner(7) requires 4 mm of lead and 10 cm of concrete. The main cause of this discrepancy between two calculations, performed for PET-CT radiation field, is the difference in adopted dose limits (see equation (2)). Most of the more recent examples from practice are executed according to the recommendations of the AAPM task group 108(3), which mean that there are not many variations. An Argentinean study(16) revealed the following differences:

COMPARISON OF VARIOUS METHODS FOR DESIGNING

Education and Science of the Republic of Serbia supported this work under contract 171007.

FUNDING The Ministry of Education and Science of the Republic of Serbia supported this work under contract 171007. REFERENCES

Page 5 of 5


1. International Atomic Energy Agency. Radiation protection in newer medical imaging techniques: PET/CT. Safety report series no. 58, Vienna (2008). 2. The Law on Protection Against Ionizing Radiation and Nuclear safety. Official Bulletin of Republic of Serbia, 36/09, Belgrade, 2009. 3. AAPM Task Group 108. PET and PET/CT shielding requirements. Med. Phys. 33(1), pp 4– 15 (2006). 4. Siemens AG Healthcare sector. Installation and System Specifications for Biograph 40/Biograph 64. Siemens AG Medical Solutions (2005). 5. Ciraj, O. and Kosˇutic´, D. Radiation Protection Program for Positron Emission Tomography and Computed Tomography (PET/CT). Institute of Nuclear Sciences (2008). 6. Din standards 6844 Part II and Part III. Deutshes Institut fu¨r Normung e.V., Alleinverkauf der Normen durch Beuth Verlag GmbH, 10772 Berlin (2006). 7. Lerner, S. Radiation protection calculation for Siemens Biograph True Point 40/64 PET CT worksheet 0940000004689-2008-NM01-001-01. Siemens AG Medical Solutions (2008). 8. Al-Haj, A. N., Lobriguito, A. M., Arafah, A. and Parker, R. Deriving staff and public doses in a PET/CT facility from measured radiation levels using thermoluminescent dosimetry. Radiat. Prot. Dosim. 144, 487 –491 (2011). 9. Mattsson, S. and So¨derberg, M. Radiation dose management in CT, SPECT/CT and PET/CT techniques. Radiat. Prot. Dosim. 147, 13– 21 (2011). 10. Castrilloń, S. V. and Henrı´quez, F. C. A study on occupational exposure in a PET/CT facility. Radiat. Prot. Dosim. 147, 247–249 (2011). 11. Lundberg, T. M., Gray, P. J. and Bartlett, M. L. Measuring and minimizing the radiation dose to nuclear medicine technologists. J. Nucl. Med. Technol. 30, 25– 30 (2002).

12. Zeff, B. W. and Yester, M. V. Patient self-attenuation and technologist dose in positron emission tomography. Med. Phys. 32, 861– 865 (2005). 13. Czarwinski, R. and Crick, M. J. Occupational exposures worldwide and revision of international standards for protection. Radiat. Prot. Dosim. 144, 2 –11 (2011). 14. Snyder, W. Estimates of absorbed fractions for monenergetic photon sources uniformly distributed in various organs of a heterogeneous phantom. J. Nucl. Med. Supplement 3, 44–45 (1969). 15. Nacional Council on Radiation Protection and Measurements. Structural shielding design and evaluation for megavoltage X- and gamma-ray radiotherapy facilities. NCRP Report No. 151 (2005). 16. Cruzate, J. and Discacciati, A. Shielding Design Considerations for PET/CT Facilities. Nuclear Regulatory Autority, Ciudad Autonoma de Buenos Aries, Republica Argentina (2008). 17. Nacional Council on Radiation Protection and Measurements. Structural Shielding Design and Evaluation for Medical Use of X Rays and Gamma Rays of Energies up to 10 MeV. NCRP Report No. 49 (1976). 18. Archer, B. R., Tomby, J. I. and Bushong, S.C. Diagnostic X-ray shielding design based on an empirical model of photon attenuation. Health Phys. 44, 507– 517 (1983). 19. Simpkin, D. J. Shielding requirements for constant-potential diagnostic x-ray beams determined by a Monte Carlo calculation. Health Phys. 56, 151– 164 (1989). 20. Marjanovic´, N., Vujisic´, M., Stankovic´, K., Despotovic´, D. and Osmokrovic´, P. Simulated exposure of titanium dioxide memristors to ion beams. Nucl. Techonol. Radiat. 25, 120– 125 (2010). 21. Metzger, R. L. and Van Riper, K. A. A Monte Carlo Shielding Model for PET/CT Clinics. Radiation Safety Engineering, Inc (20040). ´ ., Stankovic´, K., Dolicánin, D. and 22. Dolicánin, C Loncˇar, B. Statistical treatment of nuclear counting results. Nucl. Techonol. Radiat. 26, 164–170 (2011). 23. Vujisic´, M., Matijasˇevic´, D., Dolicánin, E. and Osmokrovic´, P. Simulated radiation effects in the superinsulating phase of titanium nitride films. Nucl. Techonol. Radiat. 26, 254– 260 (2011). 24. Frey, G. D. Evaluation of PET shielding. AAPM 48th Annual Meeting, Orlando (2006). 25. Martin, M. C. PET/CT—site planning and shielding design. National Symposium on Fusion Imaging and Multimodalities: Technical and Regulatory Considerations, Kansas City, 18–20 February (2004).