Dynamic PET cardiac and parametric image reconstruction: A fixed-point proximity gradient approach using patch-based DCT and tensor SVD regularization

作     者：Häggström, Ida Lin, Yizun Li, Si Krol, Andrzej Xu, Yuesheng Schmidtlein, C. Ross 

作者机构：Department of Medical Physics Memorial Sloan Kettering Cancer Center New YorkNY10065 United States School of Mathematics Guangdong Provincial Key Lab of Computational Science Sun Yat-sen University Guangzhou510275 China School of Data and Computer Science Guangdong Provincial Key Lab of Computational Science Sun Yat-sen University Guangzhou510275 China Department of Radiology Department of Pharmacology SUNY Upstate Medical University SyracuseNY13210 United States Department of Mathematics and Statistics Old Dominion University NorfolkVA23529 United States 

出 版 物：《arXiv》 (arXiv)

年 卷 期：2019年

核心收录：

主　　题：Image reconstruction 

摘      要：Our aim was to enhance visual quality and quantitative accuracy of dynamic positron emission tomography (PET) uptake images by improved image reconstruction, using sophisticated sparse penalty models that incorporate both 2D spatial+1D temporal (3DT) information. We developed two new 3DT PET reconstruction algorithms, incorporating different temporal and spatial penalties based on discrete cosine transform (DCT) w/ patches, and tensor nuclear norm (TNN) w/ patches, and compared to frame-by-frame methods;conventional 2D ordered subsets expectation maximization (OSEM) w/ post-filtering and 2D-DCT and 2D-TNN. A 3DT brain phantom with kinetic uptake (2-tissue model), and a moving 3DT cardiac/lung phantom was simulated and reconstructed. For the cardiac/lung phantom, an additional cardiac gated 2D-OSEM set was reconstructed. The structural similarity index (SSIM) and relative root mean squared error (rRMSE) relative ground truth was investigated. The image derived left ventricular (LV) volume for the cardiac/lung images was found by region growing. Parametric images of the brain phantom were calculated by nonlinear least squares fitting. For the cardiac/lung phantom, 3DT-TNN yielded optimal images with an rRMSE / SSIM of 45.4±0.4% / 0.652±0.007, compared to 65.4±0.1% / 0.4439±8E-4for conventional 2DOSEM. The optimal LV volume from the 3DT-TNN images was on average 79±4% of the true value, cardiac gated 2DOSEM 68±9%, and 2D-OSEM 24±7%. 3DT-DCT had minimum rRMSE and maximum SSIM for the brain phantom at 59.5±0.3 and 0.593±0.003% respectively, whereas 2D-OSEM had an rRMSE / SSIM of 75.6±0.4% / 0.478±0.005. Compared to 2D-OSEM, parametric images based on 3DT-DCT images had smaller bias for all six parameters, and higher SSIM for all but one. Our novel methods that incorporate both 2D spatial and 1D temporal penalties produced dynamic PET images of higher quality than conventional 2D methods, w/o need for post-filtering. Breathing and cardiac motion were simultaneously captur