Performance of quasi-exact cone-beam filtered backprojection algorithm for axially truncated helical data

Concerns the problem of tomographic reconstruction from axially truncated cone-beam projections acquired with a helical vertex path. H. Kudo et al. (1998) developed a quasi-exact filtered backprojection algorithm for this problem. This paper evaluates the performance of the quasi-exact algorithm by comparing it with variants of the approximate Feldkamp (1984) algorithm. The results show that the quasi-exact algorithm possesses the following advantages over the Feldkamp algorithm. The first advantage is the improved image quality when the pitch of the helix is large. The second advantage is the use of a detector area close to the minimum detector area that is compatible with exact reconstruction. This area is shown to be much smaller than that required by the Feldkamp algorithm for typical geometries. One disadvantage of the quasi-exact algorithm is the computation time, which is much longer than for the Feldkamp algorithm.