Huang Zhi-Feng, Kang Ke-Jun, Zhu Pei-Ping, Huang Wan-Xia, Yuan Qing-Xi, Wang Jun-Yue
Department of Engineering Physics, Tsinghua University, Beijing 100084, People's Republic of China.
Phys Med Biol. 2007 Jan 7;52(1):1-12. doi: 10.1088/0031-9155/52/1/001. Epub 2006 Dec 4.
Computed tomography of diffraction enhanced imaging (DEI-CT) is a novel x-ray phase-contrast computed tomography which is applied to inspect weakly absorbing low-Z samples. Refraction-angle images which are extracted from a series of raw DEI images measured in different positions of the rocking curve of the analyser can be regarded as projections of DEI-CT. Based on them, the distribution of refractive index decrement in the sample can be reconstructed according to the principles of CT. How to combine extraction methods and reconstruction algorithms to obtain the most accurate reconstructed results is investigated in detail in this paper. Two kinds of comparison, the comparison of different extraction methods and the comparison between "two-step" algorithms and the Hilbert filtered backprojection (HFBP) algorithm, draw the conclusion that the HFBP algorithm based on the maximum refraction-angle (MRA) method may be the best combination at present. Though all current extraction methods including the MRA method are approximate methods and cannot calculate very large refraction-angle values, the HFBP algorithm based on the MRA method is able to provide quite acceptable estimations of the distribution of refractive index decrement of the sample. The conclusion is proved by the experimental results at the Beijing Synchrotron Radiation Facility.
衍射增强成像计算机断层扫描（DEI-CT）是一种新型的X射线相衬计算机断层扫描技术，用于检测弱吸收的低Z值样品。从在分析仪摇摆曲线不同位置测量的一系列原始DEI图像中提取的折射角图像可被视为DEI-CT的投影。基于这些投影，可根据CT原理重建样品中折射率 decrement的分布。本文详细研究了如何结合提取方法和重建算法以获得最准确的重建结果。两种比较，即不同提取方法的比较以及“两步”算法与希尔伯特滤波反投影（HFBP）算法之间的比较，得出基于最大折射角（MRA）方法的HFBP算法可能是目前最佳组合的结论。尽管包括MRA方法在内的所有当前提取方法都是近似方法，无法计算非常大的折射角值，但基于MRA方法的HFBP算法能够提供相当可接受的样品折射率 decrement分布估计。在北京同步辐射装置上的实验结果证明了这一结论。
