Pinhole collimator design for nuclear survey system

Transcription

1 Annals of Nuclear Energy 29 (2002) Pinhole collimator design for nuclear survey system Wanno Lee*, Gyuseong Cho Department of Nuclear Engineering, Korea Advanced Institute of Science and Technology (KAIST), Kusong-dong, Yusong-gu, Taejon, , South Korea Received 23 August 2001; received in revised form 15 February 2002; accepted 19 February 2002 Abstract A conventional knife-edge collimator, which is widely used in gamma camera for medical diagnosis, is not suitable for nuclear imaging system because many scattering radiations near the pinhole aperture happen and blur image. A new pinhole collimator, which shapes a channeled aperture for reducing image degradation induced by the scattering radiations, is introduced and its characteristics are analyzed by Monte Carlo simulation. Resolutions defined as the full-width at half-maximum (FWHM) of point spread function and efficiencies are calculated about several pinhole diameters from 4 to 8 mm and channel heights from 2 to 10 mm. For this calculation, we assumed that 137 Cs radiation sources with 662 kev mono-energies enter into our designed collimator at the 1 m distance from the detector plane. The efficiencies and resolutions of the channeled collimator are compared with those of the conventional collimator. By comparison results, it is verified that the new collimator takes advantage more than the conventional collimator. The optimum channel height and diameter of the pinhole collimator from simulation results are also proposed and designed. We finally acquired nuclear image mounting this collimator in the nuclear survey system. # 2002 Elsevier Science Ltd. All rights reserved. 1. Introduction The collimator plays an important role determining image quality in a medical gamma camera as well as nuclear survey system. However its researches for nuclear field applications have not been accomplished unlike its enough analyses in a medical system (Smith et al., 1997; Mortimer and Anger, 1954; Johnson et al., 1995; Redus * Corresponding author. Tel.: ; fax: address: petor@cais.kaist.ac.kr (W. Lee) /02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S (02)

2 2030 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) et al., 1992). The collimator design for nuclear survey system should be differently taken into consideration from the collimator for medical environments because it has always been used in the high-energy radiation environments. Several types of collimator could be used in nuclear survey but a pinhole collimator is the most useful for nuclear imaging because large area monitoring is possible and it takes advantage of the same angular resolution irrelative to the distance between a source and a system. We have developed a nuclear survey system with the pinhole collimator that is composed of combined charge-coupled devices (CCD) and a gamma imaging system that produces a color image of gamma rays through the superimposition of white and black at surroundings of the source (Lee et al., in press). Fig. 1 shows the developed prototype system. The purpose of this study is to investigate the optimum pinhole aperture design for accomplishing the improved image at high-energy radiation field under the condition that the collimator efficiency is greater than the minimum value for being able to obtain the distinguished image from background radiation. In this paper, characteristics of the knife-edge and the channel type collimator are analyzed by the MCNP 4B photon simulation code, which can be used for neutron, photon, electron, or coupled neutron/photon/electron transport (Briesmeister, 1997). The channel type is selected by these analyses and we also verify these simulation results through experimental test of two types of collimators. After determining the collimator shape, our studies consisted of parameter calculations and Monte Carlo simulations in order to optimize the channel height and diameter. Experimental studies are used to test and evaluate the channeled pinhole collimator due to the change of two parameters. 2. Materials and methods In order to simulate interaction with radiation and detector due to aperture diameters and channel heights, the MCNP 4B transport code is used. MCNP input geometry consists of three components and has the same structure with the real senor of our developed system. The first components are the lead and aluminum parts for shielding environmental radiation, which also support the pinhole aperture. For shielding about 1.4 MeVenvironment radiations of 40 K, the lead thickness of 20 mm is selected and it can shield the half value of the 40 K. The aluminum thickness of 5 mm for supporting the lead weight is determined. The focal length defined as the distance from the pinhole aperture to scintillator crystal is mm and the field of view of the pinhole collimator is 42 when it is inserted into the system. The second component is the scintillation crystal part. We assume that the crystal type is NaI(Tl), which was made by Alpha Spectra Co. The crystal thickness, diameter, and density are respectively 10 mm, 51 mm and 3.67 g/cm 3. The gamma ray energy resolution of this crystal was proposed about 11% and the measured results showed about 12% at 662 kev.

3 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) Fig. 1. The developed prototype system: (a) radiation imaging camera without CCD visual camera is composed of the sensor, NIM bin module, and data acquisition board and (b) the acquisition software.

4 2032 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) The third is the pinhole collimator mounted on the lead head. In order to consider the experimental facility, the collimator is composed of lead cap and tungsten alloy pinhole. Fig. 2(a) shows MCNP input geometry and source distribution is shown in Fig. 2(b). The spatial resolution and efficiency are calculated by Monte Carlo method due to the collimator types and channel heights. The distance between the source and crystal plane is 1000 mm and solid angle of the source about pinhole Fig. 2. (a) Input geometry for MCNP simulation and (b) the source distribution passed the pinhole and scattered photons.

5 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) Fig. 3. Cross section of (a) knife-edge collimator and (b) channeled collimator. plane is 1.8. Geometries of the conventional and channeled pinhole are shown in Fig. 3(a) and (b). In order to evaluate the spatial resolution, we assume that the crystal is voxelized by The dimension of each voxel is 111 mm. Therefore, the maximum accuracy of resolution in this simulation is 1 mm. The collimator efficiency is calculated by summing photon numbers accumulated in the each voxel and if they are less than 5% of maximum photon numbers, they are rejected for reduction of simulation error. Collimators having several diameters and channel heights for experimental study are designed based on calculated and simulated results. For comparison of the scattering radiations created near the hole of the knife-edge and channeled collimator, we also designed the 4 mm diameter collimator of two types and measured

6 2034 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) the characteristics with standard point sources shielded by lead. As the nuclear imaging principle was described in the previous study (Lee et al., 2001), the same method was used in this paper. We compared the spatial resolutions and efficiencies of collimators having different channel heights with the line phantom and acquired the nuclear image for the channeled pinhole collimator. Fig. 4. Point spread function of (a) knife edge collimator with the 4 mm pinhole diameter and (b) channeled collimator of 2 mm height with 4 mm pinhole diameter.

7 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) Results Resolutions of the knife-edge and the channeled collimator of channel height (2 mm) with 4 mm pinhole diameter are shown in Fig. 4(a) and (b). These figures explain an effect of the channel height about image resolution. The channeled collimator can reduce image blurring because of decreasing the scattering radiation near the hole, however the image of the knife-edge collimator at high-energy radiation field broaden because of the scattered photons increased near the hole and shielding material. Therefore, the knife-edge collimator is not suitable for a nuclear survey system. According to these calculations, the spatial resolution of the channel collimator is better than that of the knife-edge by about 37.5% although the efficiency decreases by about 30.6%. These results are similar to the experimental value. Fig. 5(a) shows the FWHM of point spread function (PSF) due to the change of the channel height and diameter. As the pinhole diameter is narrow and the channel height is long, the PSF becomes sharp. Under the same conditions of Fig. 5(a), the collimator efficiencies are calculated and shown in Fig. 5(b). The efficiency inclines to decrease due to increase the channel height, as expected. In this paper the system sensitivity, which is determined by other components of an electronic system and data acquisition board besides the collimator (Guru et al., 1995; Redus et al., 1994; Gal et al., 2000), is defined as the minimum activity required for a source to be detected significantly above the noise level of the camera (Redus et al., 1996). This is described in terms of the signal-to-noise ratio as S ¼ N source N bkg pffiffiffiffiffiffiffiffiffi ¼ C source C bkg pffiffiffiffiffiffiffiffiffiffiffiffiffi C bkg =t N bkg ð1þ where S is the sensitivity, N source is the total source count detected by the nuclear survey system during the setting time, N bkg is the total background count during the same time without a source, C source is the counting rate of the source plus background, and C bkg is the counting rate of the background without the source. It is usually the minimum requirement of nuclear survey systems like our developed camera that the sensitivity should be acquired more than 6 for obtaining the meaningful image from the background noise. We used this value about the 137 Cs source of 100 Ci located in the 1000 mm distance from the detector plane during 60 s measurement time. By using Eq (1) when the background count rate (C bkg ) is 4 #/s, which was measured by our developed system, the source count rate (C source ) is calculated as a 5.55 #/s. Intrinsic and absolute efficiencies are defined as (Knoll, 1989) " int ¼ number of pulse recorded number of radiation quanta incident in detector ð2þ

8 2036 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) Fig. 5. (a) Resolutions defined as the FWHM of point spread function and (b) collimator efficiencies due to the change of channel height and diameter. number of pulse recorded " abs ¼ number of radiation quanta emitted by source : ð3þ By using Eqs. (2) and (3), the collimator efficiency is given by

9 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) " col ¼ " abs number of radiation quanta incident in detector ¼ " int number of radiation quanta emitted by source : ð4þ The peak efficiency assumes that only those interactions that deposit the full energy of the incident radiation are counted. The total and peak count efficiency are related by the peak-to-total ratio as r ¼ number of peak counts recorded by the full energy deposition number of pulses recorded ð5þ Fig. 6. The real image of line phantom shaping a cylinder that has the 10 mm length and 2 mm diameter: (a) the channel height is 2 mm and (b) the channel height is 6 mm with the 4 mm pinhole diameter.

10 2038 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) where r is the peak-to-total ratio (Knoll, 1989). If the peak cont rate with our system is measured about the known source activities, this value is then expressed by C source ¼ A " abs r ¼ A " col " int r " col ¼ C source A r " int ð6þ where A is the source activity. The count rate (C source ) and source activity (A) are proposed above. In addition, the peak-to-total ration (r) is determined as through measurements as well as simulation. The limit value level of the collimator efficiency is calculated by the Eq. (6) as 1: because the intrinsic efficiency (" int ) is actually less than 1. The limit value level based on this calculation is shown in Fig. 5(b). The collimator efficiency more than this limit value is required at the nuclear survey systems. Fig. 6(a) and (b) shows line images obtained through experiments at the same diameter (4 mm) when the channel height is changed from 2 to 6 mm. These figures apparently prove our simulation results that the increase of the channel height achieves higher resolution at the expense of decreased efficiency. Therefore, both factors should be considered simultaneously for the optimum design of the pinhole collimator. Fig. 7 shows the relative value considered the resolution and efficiency about the height and diameter having more than the limit level. The 4 mm pinhole diameter with the channel height (2 mm) is proposed as an optimum condition for the nuclear survey system. The real image using this designed Fig. 7. The relative value considered the collimator efficiencies and resolutions: it shows that the optimum value is the 4 mm pinhole diameter with the 2 mm channel height.

11 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) Fig. 8. The real radiation images at the 137 Cs point source: (a) left image is displayed by , (b) right image is displayed by pinhole collimator is given in Fig. 8 when the 137 Cs point source is emitted at the 1 m distance as explained above. 4. Conclusion The channeled pinhole collimator for application at high-energy radiation field is introduced and its characteristics are analyzed by Monte Carlo method. When we compared with resolutions of the knife-edge and the channel collimator at the same diameter, resolution degradation of the knife-edge collimator from the scattering radiation near the hole was heavier than that of the channeled collimator. Using the minimum requirements of the nuclear survey system that the sensitivity should be more than 6 for obtaining the meaningful images from background noise, the limit value of the collimator efficiency is calculated and the optimum height and diameter of pinhole is proposed. We get the real image with this designed pinhole at the 137 Cs point source. This proposed method could be applied in the optimum pinhole design for the similar nuclear survey instruments with our developed system. In the future, we will plan to obtain radiation images in the industrial fields or nuclear facilities with this pinhole. References Briesmeister, J.F., MCNP A General Monte Carlo N-Particle Transport Code Version 4B. Los Alamos National Laboratory, New Mexico.

12 2040 W. Lee, G. Cho / Annals of Nuclear Energy 29 (2002) Gal, O., Jean, F., Laine, F., Leveque, C., The CARTOGAM portable gamma imaging system. IEEE Trans. Nuclear Science 47 (3), Guru, S.V., He, Z., Wehe, D.K., Knoll, G.F., A portable gamma imaging system for radiation monitoring. IEEE Trans. Nuclear Science 42 (4), Johnson, E.L., Jaszczak, R.J., Wang, H., Li, J., Greer, K.L., Coleman, R.E., Pinhole SPECT for imaging in-111 in the head. IEEE Trans. Nuclear Science 42 (4), Knoll, G.F., Radiation Detection and Measurement. John Willey and Sons, Inc., New York. Lee, W., Cho, G., Kim, H.D. in press. A radiation monitoring system with the capability of gamma imaging and the estimation of exposure dose rate. IEEE Trans. Nuclear Science. Mortimer, R.K., Anger, H.O., The Gamma Ray Pinhole Camera with Image Amplifier. Conv. Record I.R.E., Pt 9, 2 5. Redus, R.H., Nagarkar, V., Cirignano, L.J., A nuclear survey instrument with imaging capability. IEEE Trans. Nuclear Science 39 (4), Redus, R., Squillante, M., Gordon, J., Knoll, G.F., Wehe, D., A combined video and gamma ray imaging system for robots nuclear environments. NIM in Physics Research A 353, Redus, R.H., Squillante, M., Gordon, J.S., Bennett, P., Entine, G., An imaging nuclear survey system. IEEE Trans. Nuclear Science 43 (3), Smith, F.M., Jaszczak, R.J., Wang, H., Pinhole aperture design for 131 I tumor imaging. IEEE Trans. Nuclear Science 44 (3),