Study On MicroPET And Its Image Reconstruction Algorithms

As the development of clinical diagnosis technology of PET, the microPET (micro Positron Emission Tomography system) has being developed to do research in small animal experiment. With higher spatial resolution, little cubage and lower price, it is widely used in basic medicine,disease diagnoses,medicine research and gene project,etc. This paper presents the microPET and its mathematic model, and the emphasis is on its image reconstruction arithmetic.Firstly, the development of microPET/PET is introduced. Compared with other emission imaging technologies,the significances and effects in the field of biomedicine research especially the medicament research are introduced.Secondly, according to the characteristic of the detected object of microPET/PET, appropriate mathematic model is developed;the noise of data collection is analyzed;also the key point of image reconstruction-the projection matrix P is studied, as well as the structure and calculation of matrix P are analyzed.Thirdly, the image reconstruction arithmetic of microPET/PET is studied. The representive of analytic arithmetic-Filtered Back Projection arithmetic, its principle and definition improvement by using filter function are analyzed; The normal step of iteration arithmetic is instructed, as well as the embranchments, such as Multiplication Algebraic Reconstruction Technology (MART), Maximum Likelihood Expectation Maximization (MLEM) and Ordered Subset Expectation Maximization (OSEM), are studied; OSEM is the emphasis. Its arithmetic realization,subsets partition,subsets balance and influence to the image quality and the comparisons among them are studied.Finally, 3D image reconstruction technology is introduced, direct 3D reconstruction based on 3D data collection, 3D reconstruction based on 2D data collection and 3D reconstruction by using 2D data transformed from 3D data.