Single Photon Emission Tomography Model Based On Multi Detector Elliptical Orbit And Research On Local Image Reconstruction Algorithm

As a functional imaging technology,single photon emission computed tomography(SPECT)can observe the metabolism of human body,which is of great significance for the early diagnosis of diseases.However,the traditional SPECT structure will cause some problems in the image reconstruction process,such as low signal-to-noise ratio,poor sensitivity and low image resolution.Therefore,more and more attention has been paid to overcome the defects of traditional SPECT reconstruction and generate clear and effective reconstruction images.However,at present,many scholars focus on the injection agents of myocardial perfusion imaging,rarely involving the research of SPECT equipment structure,resulting in the inherent defects of SPECT equipment has not been well solved.Therefore,it is of great practical significance to solve the inherent defects of traditional SPECT equipment and develop SPECT equipment with strong signal-to-noise ratio,high sensitivity and high image resolution.In order to solve the defect of traditional SPECT equipment in detecting heart disease,a new generation of SPECT model,E-SPECT,is proposed in this paper.This model is used for the detection of heart disease,and on this basis,corresponding simulation and algorithm research are carried out.The main work includes:(1)The structure of SPECT equipment and imaging principle are described,and the reasons of poor quality of SPECT reconstruction image are analyzed.On this basis,a new generation SPECT imaging model,E-SPECT,with elliptical orbit and multiple detectors is proposed.In addition,the computer tomography(CT)equipment is integrated for attenuation compensation and multi-modal imaging based on the E-SPECT model.(2)The ordered subsets expectation maximization(OSEM)algorithm is used to simulate the proposed E-SPECT model.First of all,it is verified from the two aspects of body contour prior knowledge and attenuation compensation that the use of CT equipment can effectively improve the quality of the reconstructed image.Secondly,the resolution of the reconstructed image is simulated,and the conclusion that the resolution of E-SPECT is 4.9mm is verified,which proves that the resolution of E-SPECT is higher than that of traditional SPECT.Finally,based on least square andmedian filtering(LSM-OSEM)algorithm is proposed to smooth the artifacts.(3)Furthermore,an ACTV-OSEM algorithm based on Total Variation regularization constraints is proposed.SPECT local reconstruction is a classical ill posed inverse problem.Although the existing iterative method can reconstruct the local image,there are serious artifacts in the image outside the reconstruction area due to the reason that the data is truncated.Minimizing the TV norm can effectively smooth the image artifacts,thereby improving the quality of the reconstruction area.In this paper,we simulate the reconstruction images of OSEM,ASD-POCS,LSM-OSEM and ACTV-OSEM by computer.After comparative analysis,we find that ACTV-OSEM algorithm can not only smooth the image effectively,but also ensure the quality of the reconstructed area.