GPU-accelerated Block Matching Algorithm for Deformable Registration of Lung CT Images.
GPU-accelerated Block Matching Algorithm for Deformable Registration of Lung CT Images.
复制标题
DOI：
10.1109/pic.2015.7489856
复制
发表时间：
2015-12
期刊：
影响因子：
--
通讯作者：
Guerrero T
中科院分区：
文献类型：
--
作者：
Li M;Xiang Z;Xiao L;Castillo E;Castillo R;Guerrero T
形变配准（Deformable Registration，DR）是医学领域的一项关键技术。然而，现有的DR方法中，许多是耗时的，配准精度需要提高，这阻碍了其临床应用。在这项研究中，我们提出了一种并行块匹配算法的肺部CT图像配准，在其中的平方差度量的总和被修改为成本函数和移动最小二乘方法被用来生成完整的位移场。该算法在统一计算设备架构（CUDA）的图形处理器（GPU）上实现。结果表明，所提出的并行块匹配方法实现了快速的运行时间，同时保持平均配准误差（标准偏差）为1.08（0.69）mm。
Deformable registration (DR) is a key technology in the medical field. However, many of the existing DR methods are time-consuming and the registration accuracy needs to be improved, which prevents their clinical applications. In this study, we propose a parallel block matching algorithm for lung CT image registration, in which the sum of squared difference metric is modified as the cost function and the moving least squares approach is used to generate the full displacement field. The algorithm is implemented on Graphic Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Results show that the proposed parallel block matching method achieves a fast runtime while maintaining an average registration error (standard deviation) of 1.08 (0.69) mm.