Research Of The Detection Angle Optimization And Geometric Correction On X-ray Fluorescence CT

With the advancement of medical science and technology,high-efficiency,lowside-effect precision treatments represented by targeted therapy have gradually become emerging solutions for cancer treatment.How to evaluate the effect of treatment is particularly important.X-ray Fluorescence CT has great application potential in this field because of its ability to simultaneously obtain imaging cross-sectional structure and trace element concentration distribution in a non-invasive manner.However,the traditional X-ray Fluorescence CT imaging equipment uses a synchrotron radiation source（SR）as source,which makes the equipment too large and expensive,which greatly hinders the development and application of X-ray Fluorescence CT.In recent years,the research and development of imaging systems based on X-ray tube sources have made desktop X-ray fluorescence CT systems become one of the hottest research topics.Further optimize the performance（detection limit）of the tube source excited X-ray fluorescence CT system and improve the reconstruction accuracy,which will help to start clinical application research earlier.In view of this,this article relies on Natural Science Foundation of Chongqing“Research on the performance improvement method and reconstruction algorithm of tube excited parallel hole collimated X-ray fluorescence CT imaging system"（project number: cstc2020jcyj-msxm X0362）to develop the X-ray fluorescence CT system Performance optimization,and research on related issues that restrict imaging quality.The main research work and research results of the thesis are:（1）Investigate the current status of X-ray Fluorescence Computed Tomography research,and track current research trends and hotspots.Systematic study carried out in terms of X-ray Fluorescence Computed Tomography mathematical principles,physical principles,imaging systems,simulation and reconstruction algorithms.（2）In view of the problem that the fluorescence signal extraction and imaging stage of X-ray Fluorescence CT system are interfered by noise,which leads to the serious artifact of reconstructed image,research and optimize the detection angle to improve system performance.A pencil-beam X-ray fluorescence CT simulation system with multiple detection angles was built by MCNP5,complete the Monte Carlo imaging simulation of GNPs solution distribution in the cylinder phantom,and FBP algorithm,SART algorithm and ML-EM algorithm were used to project data from each detection angle to reconstruct the image.By comparing and analyzing the reconstruction results,it is found that setting the fluorescence detector in the backscattering region and the vertical scattering angle can effectively improve the contrast to noise ratio of the imaging system.（3）Aiming at the problem that the reconstructed image cannot accurately reflect the information caused by the two-dimensional plane offset parameters of the imaging system.Analyze the errors caused by the offset parameters and the causes of inaccurate reconstruction of the image.Research the local linear embedding algorithm,and draw on the idea of this algorithm to design a geometric correction method based on local linear relationship.A fan beam X-ray fluorescence CT system with a built-in Nd（NO₃）₃solution phantom as the imaging sample is designed with a position shift.The data is obtained through numerical simulation experiments and Monte Carlo simulation experiments,which are used to verify the calibration method.The results show that the method can effectively correct the offset parameters.The difference between the corrected offset parameters and the preset offset parameters is very small.The reconstructed image after correction is clearer than before correction,and the element distribution is more accurate.