Document Type: Conference Proceedings
Department of radiotherapy, Faculty of Medicine, Arak University of Medical Sciences, Arak, Iran.
Department of Medical Physic, Faculty of Medicine, Tabriz University of Medical Sciences, Tabriz, Iran.
Introduction: In nuclear medicine, the use of a pixelated semiconductor detector such as CZT is an of growing interest for introducing new devices. Especially, the spatial resolution can be improved by using a pixelated parallel-hole collimator with equal holes and pixel sizes based on the pixelated detector. The purpose of this study was to compare the effect of pixelated and non-pixelated parallel hole collimators and CZT detector on the functional parameters (including the spatial resolution and sensitivity) and the image quality of a HiRe- SPECT imaging system using SIMIND Monte Carlo program.
Materials and Methods: In this study we simulated a HiRe-SPECT small animal imaging system with a pixelated CsI (Na) detector (1×1×5 mm3) and evaluated the system functional parameters and image quality. The collimator features were included; hole diameter of 1.2 mm, septal thickness of 0.2 mm, and hole length of 34 mm. We designed the pixelated collimator with rectangular hole shapes and compared with non-pixelated Hexagonal parallel hole collimator (LEHR) based on CZT detector. In pixelated collimator, hole size and pixel size were the same (1×1 mm2). Then the planar scans of a 99mTc point source and SPECT scans of a digital mini Jaszczak phantom and liver scans of a voxelized MOBY mouse phantom were simulated with SIMIND. Slices were reconstructed using the maximum-likelihood expectation maximization (MLEM) with four iterations, matrix size of 128 × 128 × 128 and voxel size of 0.3 mm. The framework of CASTOR (Version 1.0) was used for the MLEM algorithm. We assessed the spatial resolution, sensitivity and image contrast to conclude an optimized HiRe-SPECT system.
Results: The results on simulations showed that the planar spatial resolution of the pixelated and LEHR parallel hole collimators was 2.4 and 3.6 mm in terms of full width at half maximum (FWHM), respectively. Also, the sensitivity of the LEHR and pixelated rectangular collimators calculated 34.04 and 15.15 Cps/MBq, respectively. To compensate the reduction sensitivity in pixelated collimator, the CZT semiconductor detector was used. The CZT detector improved the sensitivity about 1.7 times higher than this by the CsI (Na). The minimum detectable size of hot rods in mini Jaszczak phantom on the reconstructed images were determined in the sectors of 1.8 with the pixelated collimator, and 2.4 mm for the LEHR collimator at a 5 cm distance from the phantom. Image quality and contrast of the liver of MOBY phantom and hot rods in the mini Jaszczak phantom with pixelated collimator were obtained about 1.5 times better than the LEHR collimator.
Conclusion: Based on our results, we recommend the rectangular pixelated parallel-hole collimator for a high-resolution with a CZT pixelated semiconductor detector for improving
the sensitivity of HiRe-SPECT systems.