基于异构多GPU的锥束CT图像重建研究

丛鹏, 王秉欣

丛鹏, 王秉欣. 基于异构多GPU的锥束CT图像重建研究[J]. 原子能科学技术, 2013, 47(11): 2161-2165. DOI: 10.7538/yzk.2013.47.11.2161
引用本文: 丛鹏, 王秉欣. 基于异构多GPU的锥束CT图像重建研究[J]. 原子能科学技术, 2013, 47(11): 2161-2165. DOI: 10.7538/yzk.2013.47.11.2161
CONG Peng, WANG Bing-xin. Research on Cone-beam Computed Tomographic Reconstruction Based on Heterogeneous Multi-GPU[J]. Atomic Energy Science and Technology, 2013, 47(11): 2161-2165. DOI: 10.7538/yzk.2013.47.11.2161
Citation: CONG Peng, WANG Bing-xin. Research on Cone-beam Computed Tomographic Reconstruction Based on Heterogeneous Multi-GPU[J]. Atomic Energy Science and Technology, 2013, 47(11): 2161-2165. DOI: 10.7538/yzk.2013.47.11.2161

基于异构多GPU的锥束CT图像重建研究

Research on Cone-beam Computed Tomographic Reconstruction Based on Heterogeneous Multi-GPU

  • 摘要: 针对锥束CT图像重建系统中GPU型号不一致问题,提出了基于异构多GPU的重建模型。该模型基于FDK算法进行重建,采用了按计算能力进行任务分配的方法,确保各GPU计算平衡。采用数据流分解的方法,实现了海量数据的图像重建。给出了该重建模型基于CUDA的实现方法,包括采用流管理和异步函数来实现多GPU并行计算以及滤波和反投影核函数的流程设计。利用高精度工业CT系统进行模型的实验验证。结果表明:所建立的重建模型正确有效,能充分发挥系统中异构多GPU的计算能力,执行效率高。

     

    Abstract: With respect to the problem due to the different multi-GPU types in cone-beam CT reconstruction, a model was proposed based on heterogeneous multi-GPUs. Using FDK algorithm for reconstruction, the model allocated tasks according to the computing capacity of each GPU, ensuring the balance among GPUs. Massive data image reconstruction was achieved by data flow decomposition. The implementation of the method was carried out based on CUDA, including multi-GPUs parallel computing using data flow management and asynchronous function and the design of the kernel function in filtering and back-projection. The model was tested on the high precision industrial CT system. The results illustrate that the reconstruction model is accurate and effective, taking full advantage of heterogeneous multi-GPUs, and is considerably effective compared to conventional methods.

     

  • [1] INO F, OKITSU Y, KISHI T, et al. Out-of-core cone beam reconstruction using multiple GPUs[C]∥2010 7th IEEE International Symposium on Biomedical Imaging. New York: IEEE, 2010: 792-795.
    [2] JANG B, KAELI D, DO S, et al. Multi GPU implementation of iterative tomographic reconstruction algorithms[C]∥2009 IEEE International Symposium on Biomedical Imaging. New York: IEEE, 2009: 185-188.
    [3] OKITSU Y, INO F, HAGIHARA K. High-performance cone beam reconstruction using CUDA compatible GPUs[J]. Parallel Computing, 2010, 36(2-3): 129-141.
    [4] 张舒,褚艳利,赵开勇,等. GPU高性能运算之CUDA[M]. 北京:中国水利水电出版社,2009.
    [5] 王苦愚,张定华,黄魁东,等. 一种锥束CT中平板探测器输出图像校正方法[J]. 计算机辅助设计与图形学学报,2009,21(7):954-961.WANG Kuyu, ZHANG Dinghua, HUANG Kuidong, et al. A calibrating method of flat panel detector based on cone beam computed tomography[J]. Journal of Computer-aided Design & Computer Graphics, 2009, 21(7): 954-961(in Chinese).
    [6] 石本义,王成,陈四海,等. CT断层重建中滤波函数设计的新方法[J]. CT理论与应用研究,2010,19(4):35-43.SHI Benyi, WANG Cheng, CHEN Sihai, et al. A novel method of CT reconstruction filter function design[J]. Computerized Tomography Theory and Applications, 2010, 19(4): 35-43(in Chinese).
  • 期刊类型引用(1)

    1. 韩志博,杨洪广,张建通,袁晓明,刘珊珊. 涂层及热处理对不锈钢薄壁管高温爆破性能的影响. 原子能科学技术. 2020(11): 2182-2187 . 本站查看

    其他类型引用(0)

计量
  • 文章访问数:  197
  • HTML全文浏览量:  0
  • PDF下载量:  1079
  • 被引次数: 1
出版历程
  • 刊出日期:  2013-11-19

目录

    /

    返回文章
    返回