Optimization of image quality and acquisition time for lab-based X-ray microtomography using an iterative reconstruction algorithm
Optimization of image quality and acquisition time for lab-based X-ray microtomography using an iterative reconstruction algorithm
复制标题
DOI：
10.1016/j.advwatres.2018.03.007
复制
发表时间：
2018-05-01
影响因子：
4.7
通讯作者：
Bijeljic, Branko
中科院分区：
文献类型：
--
作者：
Lin, Qingyang;Andrew, Matthew;Bijeljic, Branko
基于实验室的非侵入性 X 射线显微断层扫描已广泛应用于许多工业和研究学科。然而，与同步加速器光束线相比，使用实验室系统的主要障碍是其图像采集时间更长（每次扫描数小时，而同步加速器的数秒到分钟），这导致动态原位过程的应用受到限制。因此，现有的实验室X射线显微断层扫描大多数仅限于静态成像；相对快速的成像（每次扫描数十分钟）只能通过牺牲成像质量来实现，例如减少曝光时间或投影次数。为了缓解这一障碍，我们引入了一种众所周知的迭代重建算法的优化实现，该算法允许用户以合理的图像质量重建断层扫描图像，但与传统方法相比，需要更少的 X 射线信号计数和更少的投影。使用孔隙空间中含有和不含有液相的砂岩样品对迭代和传统滤波反投影重建算法进行定量分析和比较。总体而言，通过实施迭代重建算法，具有稀疏对象结构的样本所需的图像采集时间可减少多达 4 倍，而不会造成可测量的清晰度或信噪比损失。 (C) 2018 作者。由爱思唯尔有限公司出版
Non-invasive laboratory-based X-ray microtomography has been widely applied in many industrial and research disciplines. However, the main barrier to the use of laboratory systems compared to a synchrotron beamline is its much longer image acquisition time (hours per scan compared to seconds to minutes at a synchrotron), which results in limited application for dynamic in situ processes. Therefore, the majority of existing laboratory X-ray microtomography is limited to static imaging; relatively fast imaging (tens of minutes per scan) can only be achieved by sacrificing imaging quality, e.g. reducing exposure time or number of projections. To alleviate this barrier, we introduce an optimized implementation of a well-known iterative reconstruction algorithm that allows users to reconstruct tomographic images with reasonable image quality, but requires lower X-ray signal counts and fewer projections than conventional methods. Quantitative analysis and comparison between the iterative and the conventional filtered back-projection reconstruction algorithm was performed using a sandstone rock sample with and without liquid phases in the pore space. Overall, by implementing the iterative reconstruction algorithm, the required image acquisition time for samples such as this, with sparse object structure, can be reduced by a factor of up to 4 without measurable loss of sharpness or signal to noise ratio. (C) 2018 The Authors. Published by Elsevier Ltd.