Pet Image Reconstruction Of Key Technologies

Positron emission tomography (PET) is an integral and advanced molecular imaging technique which is a combination of photo detection, signal processing and image reconstruction, and displays the biological and physiological process by imaging the spatial distribution of radioisotope labeled chemical compounds injected into vivo. PET has become one of the most important parts of modern nuclear medicine.PET image reconstruction, as a key issue in PET research, focuses on computing the functional image from data of photo detection. Te contributions on some key problems in PET image reconstruction, are as follow:A new PET imaging model of content-adaptive point cloud is proposed in this paper, using prior information obtained from the transmission reconstruction or fast FBP reconstructed image. The primary distinction between our approach and previous efforts is the notion of image sampling: the amounts of sampling nodes are objective-variant and the generation of nodes is objective-guided in our applications. This method achieves the benefits of the accurate image representation and fast, simple computations in algorithm implementation.Another contribution of this paper is the state space representation for the PET dynamic system. A robust system identification paradigm is adopted to estimate the tracer kinetics parameters from PET sinogram data directly. It is derived and extended from the H∞filtering principles and particularly powerful for real-world situations where the modes and levels of the disturbances are unknown. Specifically, the reconstructed results gained by applying this strategy to synthetic phantom data produce persuasive assessment of our method.The availability of formulating the double-injection, single-scan protocol in PET imaging in state-space representation is also evaluated. The parallel compartmental models with kinetic parameters for dual-tracer are coupled into the state equation in biological system, and the mixed photon acquisition is integrated into measurement equation in PET reconstruction model. Finally, the results of dynamic images for each tracer are presented in simulated experiment, and the analysis on measurement sensitivity and noise sensitivity is also performed.