A geometric system model of finite aperture in small animal pinhole SPECT imaging
Introduction
Single photon emission computed tomography (SPECT) combined with a pinhole collimator can provide high resolution imaging for small laboratory animals [1], [2], [3], [4], [5]. It provides spatiotemporal information about the distribution of a radiotracer administered to a target allowing non-invasive measurement of physiological functions. Such functional in vivo imaging in animals is leading to accelerated studies of disease processes and to development and screening of drugs for diagnosis and therapy [5].
Due to the large magnification involved in imaging small objects, pinhole SPECT is sensitive to various system characteristics [6]. For example, a minor shift in center-of-rotation could lead to noticeable resolution loss [7]. Since many important SPECT isotopes, like In-111 (171, 250 keV) and I-131 (364 keV), have been applied to small animal imaging, high energy photon penetration and scattering close to pinhole aperture will also affect the spatial resolution and the sensitivity of a pinhole system because of broadening the tails of point spread function. Many recent studies have reported significant development in analytic formulation of pinhole aperture penetration [8], [9], [10], [11], [12], [13].
To avoid possible artifacts and loss of resolution, image reconstruction methods require the capability to incorporate various important physical factors of pinhole SPECT [7], [14], [15]. Statistically based iterative image reconstructions have the potential to improve on the image quality through their ability to more accurately model the underlying physics that maps the source distribution into the sinogram, as well as through their explicit modeling of the statistical variability of photon limited detection [16]. Similar research results have also demonstrated that detailed system modeling can greatly improve quantitation accuracy for both clinical and animal positron emission tomography (PET) [17], [18], [19].
Since conventional pinhole SPECT imaging assumes ideal pinhole geometry, neglect of finite aperture could lead to adverse quality degradation, such as positioning bias and image distortion. In this work, we assume that the effective size of aperture can be determined by analytic derivation or measurement. Our goal is to construct a fully 3D iterative image reconstruction for pinhole SPECT with finite aperture. Fast reconstruction algorithm using ordered subsets expectation maximization (OSEM) is implemented [20]. A system model is built to explicitly describe the process of a photon from its emission, passing through a finite aperture, to detection. Experiments using simulated phantom studies are conducted to evaluate the performance of the proposed system model compared to the system model of an ideal pinhole.
Section snippets
Statistical image reconstruction of 3D pinhole SPECT
Let λ denote an M-dimensional vector of unknown source image with each element λj representing the rate of single photon emission from the jth image voxel. Let yi denote the measured projection data along the single bin i. The emission sinogram y of 3D pinhole SPECT is an N-dimensional vector of independent Poisson-distributed photon counts with mean vector . Statistical image reconstruction requires a system model (or projection matrix) P to describe the relation of the unknown image λ to
Experiments and discussion
Consider a pinhole SPECT system consisting of a flat detector and a pinhole collimator as shown in Fig. 4. The detector plate rotates on a circular orbit during acquisition. The distance between the center of the pinhole and the origin of the xy plane is r, which is the radius of rotation (ROR). The object is placed at the center of xy plane. The center of the pinhole is on the z-axis. The distance between the center of pinhole aperture and the detector plate is f, which is the focal length of
Conclusion
In this paper, we present a system model which explicitly compensates the blurring effect due to the finite aperture in 3D pinhole SPECT for small laboratory animals. For a real pinhole SPECT, once the effective size of pinhole aperture is determined, the proposed system model is ready to incorporate this aperture size. From the simulated experiments, results indicate that a proper compensation of finite aperture can increase spatial resolution and location accuracy of the reconstructed image
Acknowledgements
This research work was supported in part by the National Science Council, Taiwan (NSC 94-2614-E-007-083, and NSC 94-2213-E-007-099), and in part by the National Health Research Institutes (NHRI ME-094-PP-05).
Ching-Han Hsu was born in Hsinchu, Taiwan, in 1966. He received the M.S. and Ph.D. degrees, both in Electrical Engineering, from the University of Southern California, Los Angles, in 1993 and 1998, respectively. From 1998 to 2000, Dr. Hsu was an assistant professor in the School of Medical Technology at the Chang Gang University, Taoyuan, Taiwan. Since 2000, he has been an assistant professor in the Department of Nuclear Science at National Tsing Hua University, Hsinchu, Taiwan. Currently, he
References (20)
- et al.
Ultra-high resolution SPECT system using four pinhole collimators for small animal studies
J Nucl Med
(1995) - et al.
Ultra high resolution pinhole SPECT for small animal study
IEEE Trans Nucl Sci
(1998) - et al.
Evaluation of high-resolution pinhole SPECT using a small rotating animal
J Nucl Med
(2001) - et al.
Design and simulation of a high-resolution stationary SPECT system for small animals
Phys Med Biol
(2004) In vivo molecular and genomic imaging: new challenges for imaging physics
Phys Med Biol
(2004)- et al.
Characterization of pinhole SPECT acquisition geometry
IEEE Trans Med Imaging
(2003) - et al.
Quantitative small field-of-view pinhole SPECT imaging: initial evaluation
IEEE Trans Nucl Sci
(1995) - et al.
Lead and tungsten pinhole inserts for i-131 SPECT tumor imaging: experimental measurements and photon transport simulations
IEEE Trans Nucl Sci
(1997) - et al.
An analytic model of pinhole aperture penetration for 3D pinhole SPECT image reconstruction
Phys Med Biol
(1998) - et al.
Analytic determination of pinhole collimator sensitivity with penetration
IEEE Trans Med Imaging
(2001)
Cited by (9)
Small-Animal SPECT, SPECT/CT, and SPECT/MRI
2021, Molecular Imaging: Principles and PracticeA Monte Carlo study on the performance evaluation of a parallel hole collimator for a HiReSPECT: A dedicated small-animal SPECT
2018, Applied Radiation and IsotopesCitation Excerpt :Also, the collimator has a significant effect on sensitivity and spatial resolution of a imaging system in a small field of view (FOV) (Azarm et al., 2015). A number of small animal SPECT imaging systems have been developed based on the pinhole or multi-pinhole collimation (Hirai et al., 2000; Hsu and Huang, 2006; Min et al., 2011; Nuyts et al., 2009; Weber et al., 1994). The studies have shown that pinholes provide lower sensitivity and detectability than parallel hole collimators (Velden et al., 2017).
Design and development of a high resolution animal SPECT scanner dedicated for rat and mouse imaging
2014, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated EquipmentCitation Excerpt :In this way, there are significant advancements happening both on the instrumentation front for data acquisition and in the development of reconstruction methods for extracting images from the acquired data. Most small animal SPECT systems have been developed based on pinhole or multi-pinhole collimation technique as reported in the literature [10–15]. Few dedicated small animal systems have been designed and developed based on parallel-hole collimation system.
Fast iterative reconstruction for helical pinhole SPECT imaging
2015, Bio-Medical Materials and EngineeringAn efficient voxel-driven system model for helical pinhole SPECT
2011, IEEE Nuclear Science Symposium Conference Record
Ching-Han Hsu was born in Hsinchu, Taiwan, in 1966. He received the M.S. and Ph.D. degrees, both in Electrical Engineering, from the University of Southern California, Los Angles, in 1993 and 1998, respectively. From 1998 to 2000, Dr. Hsu was an assistant professor in the School of Medical Technology at the Chang Gang University, Taoyuan, Taiwan. Since 2000, he has been an assistant professor in the Department of Nuclear Science at National Tsing Hua University, Hsinchu, Taiwan. Currently, he also serves as an adjunct investigator in the Division of Medical Engineering at the National Health Research Institutes, Chunan, Taiwan. His research interests include statistical image reconstruction, small animal functional imaging, and medical image processing.
Po-Chia Huang was born in Taipei City, Taiwan, in 1980. He received the B.S. and M.S. degrees, both in Nuclear Science Department from the National Tsing Hua University, Hsinchu, Taiwan in 2002 and 2004, respectively. He is currently the Research Associate at the National Tsing Hua University. His research interests include statistical image reconstruction and statistical analysis in medical images.