On Reconstruction Algorithms Of X-Ray Phase Contrast Ct By Holographic Measurements

Wenlei Ni,Ming Jiang,Shulin Zhou,Tie Zhou
2008-01-01
Abstract:Phase-contrast imaging experimentally demonstrated greatly enhanced contrast over conventional attenuation-contrast imaging for biological soft tissue. And the contrast improvements can be achieved at rather lower radiation doses. Among all of the set-ups of phase-contrast imaging, the in-line holography is the instrumentally simplest method. The mathematical model of holographic phase-contrast CT is an inverse problem of Helmholtz elliptic partial differential equation. In this paper, we discuss reconstruction algorithms for holographic X-ray phase-contrast CT. The first one is an extension of Bronnikov's work which is based on the assumption that X-ray beams travel along straight lines in the object, and Fresnel diffraction is used to describe the propagation of X-rays outgoing the object. We use the well-known Newton iterative method to solve the nonlinear relationship between the photon intensity function and the phase coefficient function. This gives a quite straightforward nonlinear phase retrieval method to reconstruct the 3D distribution of the phase coefficient of the object. The other method is an extension of the work by Jonas and Louis which is based on the Born approximations. We firstly establish a L(p) error bound for Born series expansion, and then extend the 1-order approximation reconstruction algorithm to high-order one. The results of computer simulations are discussed.
What problem does this paper attempt to address?