Computed Tomography Based On Hard X-ray Diffraction Enhanced Imaging

Diffraction Enhanced Imaging (DEI) which is one of X-ray phase-contrast imaging techniques has been one of up-to-date subjects in the X-ray imaging field. Its spatial resolution can be on the order of microns or even sub-microns by using synchrotron radiation source and it can detect organic materials and biologic soft tissue composed of light elements of the weak absorption according to the phase-contrast imaging theory, while the conventional X-ray imaging technique can only detect some matter of strong absorption and its spatial resolution is just on the order of millimeters. As so far, the exploratory researches using DEI have been done successfully in medical, biological and material fields. Thus, the research on DEI is significant both in theory and in practice.This thesis mainly does the research on the calibration of geometric misalignment in parallel computed tomography (CT) system, DEI computed tomography with multiple information and the image fusion with DEI images. The main work can be summarized as follows: Firstly, by analyzing the ideal geometric relation in parallel CT system, the thesis presents a calibration method of geometric misalignment. This method uses the relative spatial position between the axis of rotation and the detector to describe the geometric misalignment of the whole CT system. According to the imaging character of parallel CT, the method adjusts two parameters—angle offset and displacement offset to calibrate the geometric misalignment. The reconstruction results are much better after the calibration.Secondly, DEI-CT is researched as the core of the thesis. The absorption information, extinction information and refraction information extracted from DEI images are respectively used to reconstruct the object. The thesis discusses the feasibility of combining CT techniques with the information mentioned above and places a strong emphasis on refraction CT. According to the rotation invariability, the refraction CT can be classified into two modes: invariance-rotation refraction CT and variance-rotation refraction CT. In invariance-rotation refraction CT, the thesis indicates that the existing CT algorithms can be used to reconstruct the object directly. In variance-rotation refraction CT, the adjusting-projection method is presented for reconstruction by analyzing the property of the gradient rotating. At the same time, another method — inverse Radon transform is presented to be used in variance-rotation refraction CT since the refraction information is the integral of first order derivative of...