Study On Cone Beam CT Reconstruction Algorithm And Its Application

Computed tomography (CT) is a kind of technology that widely used in medical,industrial, military and aerospace. With the emergence of flat-panel detector, cone beam CTreconstruction becomes the focus of current research. Compared with parallel beam and fanbeam reconstruction, cone beam CT reconstruction has the advantage that faster projectiondata acquisition, higher ray utilization and better axial resolution. In this paper, study on conebeam CT reconstruction algorithm and its application. The main research content is asfollows:1. Based on the study of the two-dimensional fan beam reconstruction algorithm, thefilter function in filtering process is improved. A new kind of filter function, which combinedwith the idea of weighted average and the hybrid filter, is put forward. Compared withtraditional filter function, the reconstruction image using this filter function can effectivelyinhibit oscillation and make reconstruction image smoother. In addition, its noise performanceis higher than the traditional filter functions.2. Study on cone beam FDK algorithm, from the perspective of data sufficiencycondition and3D Radon transform doing the artifact analysis for FDK algorithm. Study theacceleration of every step of FDK algorithm. The new filter function is applied to the FDKalgorithm. The same good reconstruction results are obtained.3. Study on the ring artifact correction in cone beam CT reconstruction First of all,analyzes the cause, characteristics and effects of ring artifact. Then, improved a kind of ringartifact correction methods that after image reconstruction. Compared with the originalmethod, it can better correct the ring artifacts and keep the image details. At last, aim atmultiple consecutive detector damage, put forward in a sinogram correction method whichcombined with the LoG detection and Hough transform. By comparison, this method can remove that type ring artifacts effectively.