Anthropomorphic radiation therapy phantoms: a quantitative assessment of tissue substitutes.

Compositional analyses of samples of soft tissue and lung substitutes used in the construction of anthropomorphic radiotherapy phantoms were performed for products from Alderson Research Laboratories Inc., Kyoto Scientific Specimens Company/Capintec Inc., Humanoid Systems Inc., Radiology Support Devices Inc., and The Phantom Laboratory Inc. To assess quality control in the manufacture process, the results of these compositional analyses were compared with individual manufacturer's design specifications and electron densities measured by computed tomography analysis of actual phantom slices. A comparison of the manufacturers design specifications against reference tissue parameters from Tissue Substitutes in Radiation Dosimetry and Measurement (International Commission on Radiation Units and Measurements) indicates marked differences in the basic design of commercially available anthropomorphic phantoms.

Assessment of 18F gaseous releases during the production of 18F-fluorodeoxyglucose.

Fluorodeoxyglucose labeled with 18F (18F-FDG) is the most commonly used radiopharmaceutical in positron emission tomography (PET). Fluorine-18-labeled FDG is used as a diagnostic tool in PET studies to monitor the physiology of the brain, diagnose heart function and disease, and to image cancerous tumors. At the University of California, Los Angeles (UCLA), three cyclotrons produce [18F]-fluoride ion using 18O-enriched water targets. Fluorine-18, which has a half-life of 109.8 min, is produced using an 18O(p.n.)18F reaction and is chemically processed to yield 18F-FDG. This study presents data which demonstrate that during the radiochemical processes involved in the production of 18F-FDG, gaseous effluent containing 18F is released. Forty cyclotron production runs with average end of cyclotron bombardment activities of 15.9 +/- 1.88 GBq (430 +/- 50.8 mCi) and end of radiochemical synthesis activities of 5.40 +/- 1.27 GBq (146 +/- 34.3 mCi) yield 18F gaseous effluent releases ranging from 0 to 2560 MBq (0 to 69.2 mCi) with a mean of 437 MBq (11.8 mCi). Temporal correlation of the 18F gaseous releases during 18F-FDG radiochemical production has tied the 18F release to the addition of the glucose precursor (mannotriflate) and ethyl ether in the radiochemical processing. The results are presented in terms of activities released and dilution factors required from the release stack point to maintain controlled (occupational) and uncontrolled (public) area limits in accordance with the recommendations of the International Commission on Radiological Protection and the regulatory requirements of the federal government.

Measurements of accelerator-produced leakage neutron and photon transmission through concrete.

Optimum shielding of the radiation from particle accelerators requires knowledge of the attenuation characteristics of the shielding material. The most common material for shielding this radiation is concrete, which can be made using various materials of different densities as aggregates. These different concrete mixes can have very different attenuation characteristics. Information about the attenuation of leakage photons and neutrons in ordinary and heavy concrete is, however, very limited. To increase our knowledge and understanding of the radiation attenuation in concrete of various compositions, we have performed measurements of the transmission of leakage radiation, photons and neutrons, from a Varian Clinac 2100C medical linear accelerator operating at maximum electron energies of 6 and 18 MeV. We have also calculated, using Monte Carlo techniques, the leakage neutron spectra and its transmission through concrete. The results of these measurements and calculations extend the information currently available for designing shielding for medical electron accelerators. Photon transmission characteristics depend more on the manufacturer of the concrete than on the atomic composition. A possible cause for this effect is a non-uniform distribution of the high-density aggregate, typically iron, in the concrete matrix. Errors in estimated transmission of photons can exceed a factor of three, depending on barrier thickness, if attenuation in high-density concrete is simply scaled from that of normal density concrete. We found that neutron transmission through the high-density concretes can be estimated most reasonably and conservatively by using the linear tenth-value layer of normal concrete if specific values of the tenth-value layer of the high-density concrete are not known. The reason for this is that the neutron transmission depends primarily on the hydrogen content of the concrete, which does not significantly depend on concrete density. Errors of factors of two to more than ten, depending on barrier thickness, in the estimated transmission of neutrons through high-density concrete can be made if the attenuation is scaled by density from normal concrete.

WENDI: an improved neutron rem meter.

Neutron rem meters are routinely used for real-time field measurements of neutron dose equivalent where neutron spectra are unknown or poorly characterized. These meters are designed so that their response per unit fluence approximates an appropriate fluence-to-dose conversion function. Typically, a polyethylene moderator assembly surrounds a thermal neutron detector, such as a BF3 counter tube. Internal absorbers may also be used to further fine-tune the detector response to the shape of the desired fluence conversion function. Historical designs suffer from a number of limitations. Accuracy for some designs is poor at intermediate energies (50 keV-250 keV) critical for nuclear power plant dosimetry. The well-known Andersson-Braun design suffers from angular dependence because of its lack of spherical symmetry. Furthermore, all models using a pure polyethylene moderator have no useful high-energy response, which makes them inaccurate around high-energy accelerator facilities. This paper describes two new neutron rem meter designs with improved accuracy over the energy range from thermal to 5 GeV. The Wide Energy Neutron Detection Instrument (WENDI) makes use of both neutron generation and absorption to contour the detector response function. Tungsten or tungsten carbide (WC) powder is added to a polyethylene moderator with the expressed purpose of generating spallation neutrons in tungsten nuclei and thus enhance the high-energy response of the meter beyond 8 MeV. Tungsten's absorption resonance structure below several keV was also found to be useful in contouring the meter's response function. The WENDI rem meters were designed and optimized using the Los Alamos Monte Carlo codes MCNP, MCNPX, and LAHET. A first generation prototype (WENDI-I) was built in 1995 and its testing was completed in 1996. This design placed a BF3 counter in the center of a spherical moderator assembly, whose outer shell consisted of 30% by weight WC in a matrix of polyethylene. A borated silicone rubber (5% boron by weight) absorber covered an inner polyethylene sphere to control the meter's response at intermediate energies. A second generation design (WENDI-II) was finalized and tested in 1999. It further extended the high-energy response beyond 20 MeV, increased sensitivity, and greatly facilitated the manufacturing process. A 3He counter tube is located in the center of a cylindrical polyethylene moderator assembly. Tungsten powder surrounds the counter tube at an inner radius of 4 cm and performs the double duty of neutron generation above 8 MeV and absorption below several keV. WENDI-II is suitable for field use as a portable rem meter in a variety of work place environments, and has been recently commercialized under license by Eberline Instruments, Inc. and Ludlum Measurements, Inc. Sensitivity is about a factor of 12 higher than that of the Hankins Modified Sphere (Eberline NRD meter) in a bare 252Cf field. Additionally, the energy response for WENDI-II closely follows the contour of the Ambient Dose Equivalent per unit fluence function [H'(10)/phi] above 0.1 MeV. Its energy response at 500 MeV is approximately 15 times higher than that of the Hankins and Andersson-Braun meters. Measurements of the energy and directional response of the improved meter are presented and the measured response function is shown to agree closely with the predictions of the Monte Carlo simulations in the range from 0.144 MeV to 19 MeV.

Neutron sources in the Varian Clinac 2100C/2300C medical accelerator calculated by the EGS4 code.

The photoneutron yields produced in different components of the medical accelerator heads evaluated in these studies (24-MV Clinac 2500 and a Clinac 2100C/2300C running in the 10-MV, 15-MV, 18-MV and 20-MV modes) were calculated by the EGS4 Monte Carlo code using a modified version of the Combinatorial Geometry of MORSE-CG. Actual component dimensions and materials (i.e., targets, collimators, flattening filters, jaws and shielding for specific accelerator heads) were used in the geometric simulations. Calculated relative neutron yields in different components of a 24-MV Clinac 2500 were compared with the published measured data, and were found to agree to within +/-30%. Total neutron yields produced in the Clinac 2100/2300, as a function of primary electron energy and field size, are presented. A simplified Clinac 2100/2300C geometry is presented to calculate neutron yields, which were compared with those calculated by using the fully-described geometry.

Neutron fluence and energy spectra around the Varian Clinac 2100C/2300C medical accelerator.

We have simulated the head geometry of a Varian Clinac 2100C/2300C medical accelerator in a Monte Carlo calculation to produce photoneutrons and transport them through the head shielding into a typical therapy room (modeled by a test cell at Varian Associates). The fast neutron leakage fluence and energy spectra have been calculated at 7 positions around the linac head for typical beam operation at 10, 15, 18 and 20 MV. The results of these calculations have been compared with limited measurements made using the same model accelerator operating in a Varian test cell. Calculations were also made for the fluence and energy spectra outside the head with no surrounding concrete walls, floor or ceiling to eliminate the effects of scattering from concrete. Comparisons were also made with calculations using a much simplified head geometry. The results indicate that the calculations using the complex head geometry compare, within the uncertainties, with the measurements. The simple head geometry leads to differences of a factor of 2 from the complex geometry. Results of these calculations can be used to calculate fast neutron transmission through various shielding configurations and through labyrinths.