Generation And Performance Analysis Of The System Response Matrix In Dual-Head PET Based On The Ray-tracing Algorithm

With the development of biological technology, medical imaging technology has also been improved. As one of the most important technologies in the field of nuclear medicine, Positron emission tomography plays a huge role in clinical diagnosis. The system response matrix of the image reflects the mathematical relationship between the source and the projection data, so the quality of the system has a decisive influence on the quality of reconstructed image. All in all, the computation of the system matrix is the most important step in PET image reconstruction.In this thesis, how to improve the ray-tracing algorithm was studied by using the traditional ray-tracing algorithm and the characteristics of the PET system. A simple algorithm based on line segment weighted was proposed and was optimized by subdividing. Then a new algorithm based on ray-tracing algorithm and Monte-Carlo method algorithm was proposed at the end of this thesis.This thesis is mainly divided into the following parts:(1) Introducing the principle of the ray tracing algorithm and several optimization algorithms. A ray tracing optimization algorithm based on line segment weighted is realized and a method of storing a system response matrix is designed. The image reconstruction of the system response matrix is carried out by using the rectangular and point phantom.(2) Two algorithms were used to optimize the ray tracing algorithm, which were based on the subdivision of the voxel and the subdivision of the crystal detector respectively. In the analysis found that only voxel subdivision of the reconstructed image would appear “particle" problem. In order to solve the problem of "particles" and improve the resolution, we make analyses of system matrix based on the crystal subdivision. In order to verify the performance of the system matrix obtained by different methods, the corresponding simulation experiments and physical phantom experiments are carried out. The experimental results show that the system matrix obtained by the method of crystal subdivision has better reconstruction effect compared with the method that does not do subdivision and the simple method of pixel subdivision.(3) A new algorithm is proposed to improve the quality of the system response matrix of ray tracing algorithm. Firstly, the position of the voxel is located by the ray tracing algorithm. Then the point spread function of the response line based on Monte Carlo method is extracted. Finally, the corresponding system response matrix is obtained by the convolution operation of the point spread function. The experimental results show that the system response matrix of the system is similar to that of the Monte Carlo method, and improves the quality of the system response matrix of the original ray tracing algorithm.