| Computed tomography(CT) has been a great development during decades. The scan imaging modalities have changed from a two-dimensional tomographic imaging to three-dimensional cone-beam imaging. Due to the high scanning speed, high resolution, and excellent image quality, three-dimensional cone-beam CT has been widely applied to application. In addition, it is one of the CT technology development directions. The three-dimensional cone-beam CT system has extremely stringent requirements on the ray source, detector and object to be scanned for the geometry and mounting accuracy, which greatly improves the difficulty of installation and commission of the system manufacturer. This thesis does research on the measurement of cone beam CT system geometric parameters and correction problems. The aim is to solve the system geometry parameters automatically and correction effectively. The main work and innovations are as follows:Firstly, analyzes the geometry requirements of the FDK image reconstruction algorithm commonly used in cone-beam CT theoretically. What's more, analyze the differences between the ideal cone-beam CT system and the actual cone-beam CT system as well as the influence of geometric parameters on the reconstructed image by simulation, if there are geometric errors of the cone-beam CT system. Furthermore, analyzes the degree of reconstructed image influenced by different geometric parameters under different cone angle or different magnifications conditions,Secondly, proposed a calibration correction method of geometrical parameters based on the line phantom. Design a dedicated calibration phantom to obtain the geometric parameters. Theoretically analyze and derive the geometry relationships between the calibration phantom and the projection data, and obtain the equations solving system geometric parameters. Obtain the geometric parameters of the system before the reconstruction according to the relationships between the parameters and the projection data. Simulation and actual experiments results show the line phantom achieves higher accuracy of solving geometry parameters compared with point phantom, and it can resist noise well.Thirdly, reconstruct image using modified FDK algorithm, which has well calibrated the geometric errors of cone beam CT system. Verify the correctness of calibration method base on line phantom by simulation. The experiments demonstrate that the proposed method has good noise immunity, and the positioning accuracy can meet the needs of practical application. What's more, compared with calibration method based on the point-like phantom, the results using line phantom to correct the geometric parameters are better. |
PDF Full Text Request