Premium
SPECT brain imaging of the dopaminergic system in Parkinsonism using I 123 and Tc 99 m labeled agents
Author(s) -
Du Yong
Publication year - 2004
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.1820014
Subject(s) - collimator , spect imaging , computer science , single photon emission computed tomography , image quality , iterative reconstruction , monte carlo method , medical imaging , partial volume , image resolution , detector , computer vision , artificial intelligence , nuclear medicine , physics , optics , mathematics , medicine , image (mathematics) , telecommunications , statistics
SPECT brain imaging of the dopaminergic system using I123 and Tc99 mlabeled agents, especially the simultaneous imaging of both pre‐ and postsynaptic neurons, promises to provide accurate diagnosis and differentiation of Parkinsonism. However, there are many degrading factors that affect the quality and quantitative accuracy of the SPECT images. These degrading factors limit the potential clinical applications of brain SPECT imaging. In this work, we studied these degrading factors by developing and validating a Monte Carlo (MC) method that provides accurate SPECT simulation with detailed modeling of the photon interactions inside the collimator detector system. To compensate for the partial volume effect (PVE) in the SPECT images caused by finite spatial resolution, we developed a new PVE compensation method that takes into account the effects of nonlinearity in iterative reconstruction‐based compensation for image degrading factors, including attenuation, scatter, and collimator detector response. Compensation using the new method greatly improved the quantitative accuracy of brain SPECT images. We have also developed model‐based method that can accurately estimate the downscatter and crosstalk contamination in the I123 imaging and the simultaneousI123 / Tc99 mdual‐isotope imaging. Based on the model‐based method, two different approaches to model‐based downscatter and crosstalk contamination compensation were proposed. Both methods are based on iterative reconstruction and include compensation for other imaging degrading factors. The model‐based downscatter and crosstalk compensation method provided greatly improved accuracy of activity estimates with little effect on the precision. Finally, optimization of energy windows for simultaneousI123 / Tc99 macquisition was performed to find the energy windows with the best trade‐off between minimizing the crosstalk and maximizing the detection efficiency for simultaneous acquisitions. In summary, comprehensive methods were developed and evaluated to compensate for image degrading factors in simultaneous dual‐isotope brain SPECT imaging. Application of these methods in the imaging of the dopaminergic system has the potential to provide improved accuracy for diagnosis of Parkinsonism.