Improvement based on FDK reconstruction algorithm of conebeam CT

As a typical 3D image algorithm, Feldkamp-David-Kress reconstruction algorithm is a filtered backprojection very similar to the 2D algorithm. Recent years, it has been widely used for its easy implementation and acceptable reconstruction precision for small cone-beam angle. For big cone-beam angles, however, the reconstruction exactness of horizontal planes will dramatically decrease with the distance from the central plane increases. Therefore, for improving the reconstruction precision in vertical direction, a similar reciprocal Gaussian function as the weighted factor was introduced into FDK algorithm in the article. The validity of the improved FDK algorithm was verified and evaluated through both the computer numerical simulation and the phantom model experiment. The quantitative analysis for the reconstruction results demonstrated that the reconstruction image using the improved FDK algorithm could primely revise the original reconstruction images and restore well closely to the tomography image of the prototype object.