Calibration and Imaging Modeling of CT Systems Based on Least Squares Fitting and Filtered Inverse Projection Algorithms

This study is devoted to the design and application of algorithms based on planar CT technology. First, we propose a set of accurate calibration algorithms for the CT system, and successfully determine the position of the center of rotation, the inter-detector cell distance, and the direction of rotation of the X-rays through the processing of the projection data of the calibration plate by the least-squares fitting and the optimization of the parameters. Secondly, we applied the calibration parameters to the imaging reconstruction of unknown media and combined the Filtered Back Projection algorithm (FBP) and the Algebraic Reconstruction Technique (ART) to establish two kinds of CT back projection reconstruction models, continuous and discrete, to realize the accurate determination of the media position, geometry, and absorptivity. And meanwhile, it can provide the absorptivity information at the specified position. Further, we analyzed the accuracy and stability of the calibration algorithm and verified the performance of the algorithm under different noise conditions through simulation experiments, and the results showed that the algorithm has high accuracy and stability. Overall, this study has made significant progress in CT system calibration and imaging reconstruction, providing reliable support and guidance for the application of CT technology in medicine and engineering.