Department of Engineering Physics, Tsinghua University, Beijing, China.
J Xray Sci Technol. 2011;19(3):403-15. doi: 10.3233/XST-2011-0303.
Differential phase contrast imaging computed tomography (DPCI-CT) is a novel x-ray inspection method to reconstruct the distribution of refraction index rather than the attenuation coefficient in weakly absorbing samples. In this paper, we propose an iterative reconstruction algorithm for DPCI-CT which benefits from the new compressed sensing theory. We first realize a differential algebraic reconstruction technique (DART) by discretizing the projection process of the differential phase contrast imaging into a linear partial derivative matrix. In this way the compressed sensing reconstruction problem of DPCI reconstruction can be transformed to a resolved problem in the transmission imaging CT. Our algorithm has the potential to reconstruct the refraction index distribution of the sample from highly undersampled projection data. Thus it can significantly reduce the dose and inspection time. The proposed algorithm has been validated by numerical simulations and actual experiments.
差分相衬成像计算机断层扫描（DPCI-CT）是一种新型的 X 射线检测方法，用于重建折射指数的分布，而不是弱吸收样品中的衰减系数。在本文中，我们提出了一种基于压缩感知理论的迭代重建算法。我们首先通过将差分相衬成像的投影过程离散化为线性偏导数矩阵，实现了一种差分代数重建技术（DART）。这样，差分相衬成像重建的压缩感知重建问题就可以转化为传输成像 CT 中的一个可解问题。我们的算法有可能从高度欠采样的投影数据中重建样品的折射指数分布。因此，可以显著减少剂量和检查时间。该算法已经通过数值模拟和实际实验进行了验证。
