A Prototype High Resolution MPPC-based PET System For Small Animal Imaging

Positron emission tomography (PET) system holds one of the most important applications in nuclear medical imaging device as a biomedical research technique, drug development and clinical diagnostic procedure. The small animal PET is one of the most important branches of the PET system, and it plays an essential role in the preclinical research on pharmacology and anatomy. The fast development of the pharmacy and healthcare services results in an urgent demand for the small PET system with higher performance in recent years. Therefore, many researchers have paid attention to developing a high performance small animal PET system.In this paper, we have designed a prototype of small animal PET system based on MPPC. And we also have proposed a novel dual-tracer PET imaging method using Cerenkov light. The main content and contributions are as followed:First of all, we have designed a novel PET detector consisting of a Hamamtsu 8 X 8 MPPC array module coupled to a 16 X 16 LYSO scintillator array. The size of the each scintillator is 1.5×1.5×15 mm3. The results from both Monte Carlo simualtion and tests on the designed detector show that the energy resolution of the detector is 23.6% and the spatial resolution is 1.3 mm. For the timing resolution of the PET detector, we have proposed two novel timing calibration methods, Iterative shrinkage thresholding(IST) method and Total variation merge (TV merge) method. In these two mehtods, the timing calibration process is formulated into a linear problem. To achieve a high-accurate timing calibration, a L1 norm constraint and a TV constraint are added in this linear equation, respectively. The proposed methods can achieve a true-crystal level timing calibration compared to traditional methods and results in a large improvement of the timing resolution fo the detector and system.Secondly, we propose a novel low rank-based activity map reconstruction scheme from emission sinograms of dynamic PET, termed as SLCR representing sparse/low rank constrained reconstruction for dynamic PET Imaging. The linearized alternating direction method is used to get the solution of the objective function for dynamic PET imaging. The proposed SLCR method can achieve a high quality reconstruction without specific assumption on the statistical model of data, especially when the count of the data is low. Based on the Monte Carlo simulation and real scanning data, the bias, variance and contrast recovery coefficient of the results of the SLCR method is improved by 3% to 9% compared to the tradition methods. In addition, based on the low rank constraint and sparse representation, the dynamic information and stationary information can be separated during the reconstruction. Therefore, the proposed method can enhance the contrast of the target region and provide more information for clinical and diagnosis requirements. Furthermore, to accelerate the image construction and make it more useful for application, we have developed a PET reconstruction platform using the GPU (Graphics Processing Unit) acceleration. With the help of the GPU acceleration, we can obtain the high-quality images quickly.Finally, we have designed a MPPC-based samll animal PET system. It consists of two ringunit, and there are 8 detector blocks in each ringunit. The trans-axial field of view (FOV) of system is 65 mm and the axial FOV is 70 mm. Based on the results of the Monte Carlo simulation, the sensitivity of the system is 0.5%, the spatial resolution is 1.6 mm and the sactter fraction is 19.7%. In the dual-tracer PET imaging, at the first time, the Cerenkov light, which is generated during the positron emission process of the nuclide, is used as the unique mark for different tracers. We can use the different ligh spectrum of the Cerenkov light to realize a dual-tracer PET imaging. It solves the signal separation problem of the dual-tracer imaging. And the results of the Monte Carlo simulation have verified the feasibility of the proposed method.