Image reconstruction algorithm for in-line phase contrast imaging computed tomography with an improved anisotropic diffusion model

Abstract

Phase contrast imaging (PCI) is a new physical and biochemical technique. Practical biomedical applications combine PCI with computer tomography (CT), Phase contrast CT (PC-CT) can reconstruct 3D images of samples. How to reconstruct high quality image at a low radiation dose level is a hot topic for PC-CT. In order to reduce radiation dose, a strategy is to collect incomplete projection data by few-view projection data. This work presents a reconstruction method for incomplete data PC-CT. It is based on an algebraic iteration reconstruction algorithm and combined with an anisotropic diffusion model to reduce aliasing distortions.

To validate the availability of this method, the research carried out a computer-simulated and real experimental synchrotron data. The computer-simulated and real data results demonstrate that the presented method can improve the convergence speed of image reconstruction and reduce the aliasing distortions by incomplete projection data for PC-CT. However, there is no proof that this is true for all kinds of structures.