Research On Filtered Back Projection Algorithm In 3D Cone-beam CT

In recent years, the three-dimensional cone-beam CT have more and more widespread application in the medicine and the industry, simultaneously the cone-beam reconstruction algorithm also have the rapidly expand. Among all kinds of reconstruction algorithms for 3D cone-beam CT, Filtered back projection algorithm is a type most commonly used reconstruction algorithm, so it is very important in medical and non-destruction.base on the some filtered back projection algorithm in three-dimensional cone-beam CT, in order to improve the image reconstructed and restrain its noise. We did some studies and explorations about The FDK algorithm and the Katsevich algorithm in this paper.First of all, we conducted some research in the filter function and the interpolating function. First of all, we drawn out the FDK algorithm nature and confirmed these nature by the simulation result and mathematics inferential reasoning. Then, in the filter function aspect, we put forward a mixed filter function, which are integration of the R-L and S-L filer function. From the results of the actual reconstruction, we know that the results from this kind filter function have a higher Space resolution in both horizontal and vertical, and in the spatial and temporal resolution, it has better results. Then, in interpolating function aspect, In the back-projection process of the FDK algorithm,every voxel's project distribution in the detector have itself characteristics,we presents a new interpolation method, and actual experimental results show that the FDK algorithm use this new interpolation method can restrain noise effect.At last, this paper also has carried on certain analysis and the research to the helical scan path's precise reconstruction algorithm - - Katsevich algorithm's formula and the step(including derivatives of cone-beam data, filtration of derivative data and backprojection of last data),and has done the simulation experiment.It has built the foundation for the later related work. .Key word:FDK algorithm, Nature, Mixed filter function, Interpolation, Katsevich algorithm...