Comparative Analysis of Reconstruction Algorithms for Multi-slice Clinical CT

Currently, there are major Feldkamp-Davis-Kress (FDK) based three dimensional filtered back projection (FBP) algorithms and fan-to-parallel rebinning followed by two dimensional FBP algorithms for multi-slice clinical CT reconstruction. This paper compared four different reconstruction algorithms, including conventional FDK algorithm, two-dimensional-weighted cone beam filtered back projection (CB-FBP) algorithm, three-dimensional-weighted CB-FBP algorithm and adaptive axial interpolator (AAI-FBP) algorithm. We performed a experiment on simulated helical projection data of Shepp-Logan phantom. After that, a detailed comparative performance analysis of the four algorithms is given. Conclusion:we find that AAI-FBP might be the optimal choice for conventional multi-slice clinical CT reconstruction. We hope this paper could be helpful for the improvement of multi-slice CT reconstruction algorithm as well as clinical application research.