Microlocal analysis of an FBP algorithm for truncated spiral cone beam data

    Authors

    A. Katsevich

    Comments

    Authors: contact us about adding a copy of your work at STARS@ucf.edu

    Abbreviated Journal Title

    J. Fourier Anal. Appl.

    Keywords

    cone-beam; spiral tomography; approximate reconstruction; filtered; back-projection; algorithm; analysis of artifacts; FILTERED-BACKPROJECTION ALGORITHM; X-RAY TRANSFORM; HELICAL DATA; LONG-OBJECT; TOMOGRAPHY; Mathematics, Applied

    Abstract

    In this article we propose an FBP-type algorithm for inversion of spiral cone beam data, study its theoretical properties, and illustrate performance of the algorithm by numerical examples. In particular it is shown that the algorithm does not reconstruct f exactly, but computes the result of applying a pseudo-differential operator (PDO) with singular symbol to f. Away from critical directions the amplitude of this PDO is homogeneous of order zero in the dual variable, bounded, and approaches one as the pitch of the spiral goes to zero. Numerical experiments presented in the article show that even when the pitch is relatively large, the accuracy of reconstruction is quite high. On the other hand, under certain circumstances, the algorithm produces artifacts typical of all FBP-type algorithms.

    Journal Title

    Journal of Fourier Analysis and Applications

    Volume

    8

    Issue/Number

    5

    Publication Date

    1-1-2002

    Document Type

    Article

    Language

    English

    First Page

    407

    Last Page

    425

    WOS Identifier

    WOS:000177517000001

    ISSN

    1069-5869

    Share

    COinS