A distributed SIRT implementation for the ASTRA Toolbox

A distributed SIRT implementation for the ASTRA Toolbox
复制标题

ASTRA Toolbox 的分布式 SIRT 实现

DOI:
--
复制
发表时间:
2015
期刊:
影响因子:
--
通讯作者:
K. Mueller
K. Mueller
中科院分区:
--
文献类型:
--
作者:
W. J. Palenstijn;J. Bédorf;J. Batenburg;M. King;S. Glick;K. Mueller

文献摘要

被引文献

相似文献

textabstract ASTRA ®是一个软件工具箱,可以快速开发GPU加速 层析成像算法它包含正向和反向投影操作的GPU实现 用于常见的扫描几何形状,以及一组用于迭代重建的算法。这些 算法当前限于使用单个GPU。 迭代的一个缺点 重建算法的一个缺点是,与经典的重建算法相比, 反投影算法因此,仅使用单个GPU可能会导致 在处理大数据量时,重建时间过长。 在本文中,我们提出了ASTRA工具箱的扩展,并实现了向前 投影,反投影和SIRT算法,可以分布在 多个GPU和多个工作站,以处理更大的数据集 使用ASTRA可行。
textabstractThe ASTRA Toolbox is a software toolbox that enables rapid development of GPU accelerated tomography algorithms. It contains GPU implementations of forward and backprojection operations for common scanning geometries, as well as a set of algorithms for iterative reconstruction. These algorithms are currently limited to using a single GPU. A drawback of iterative reconstruction algorithms is that they are slow compared to classical backprojection algorithms. As a result, using only a single GPU can result in prohibitively long reconstruction times when working with large data volumes. In this paper, we present an extension of the ASTRA Toolbox with implementations of forward projection, backprojection and the SIRT algorithm that can be distributed over multiple GPUs and multiple workstations to make processing larger data sets with ASTRA feasible.