Fast reconstruction of 3D volumes from 2D CT projection data with GPUs

BMC Research Notes

James Brock²  Saoni Mukherjee¹  Miriam Leeser¹ 

[1] Department of Electrical and Computer Engineering, 440 Dana Building, Northeastern University, 360 Huntington Ave, Boston, MA 02115, USA;Cognitive Electronics, 201 South St, Suite 301, Boston, MA 02111, USA

关键词: OpenCL;    CUDA;    Conebeam reconstruction;    Graphics processing unit;    Computed tomography;   

Others  :  1130058
DOI  :  10.1186/1756-0500-7-582

 received in 2014-03-03, accepted in 2014-08-18,  发布年份 2014

【 摘 要 】

Background

Biomedical image reconstruction applications require producing high fidelity images in or close to real-time. We have implemented reconstruction of three dimensional conebeam computed tomography(CBCT) with two dimensional projections. The algorithm takes slices of the target, weights and filters them to backproject the data, then creates the final 3D volume. We have implemented the algorithm using several hardware and software approaches and taken advantage of different types of parallelism in modern processors. The two hardware platforms used are a Central Processing Unit (CPU) and a heterogeneous system with a combination of CPU and GPU. On the CPU we implement serial MATLAB, parallel MATLAB, C and parallel C with OpenMP extensions. These codes are compared against the heterogeneous versions written in CUDA-C and OpenCL.

Findings

Our results show that GPUs are particularly well suited to accelerating CBCT. Relative performance was evaluated on a mathematical phantom as well as on mouse data. Speedups of up to 200x are observed by using an AMD GPU compared to a parallel version in C with OpenMP constructs.

Conclusions

In this paper, we have implemented the Feldkamp-Davis-Kress algorithm, compatible with Fessler’s image reconstruction toolbox and tested it on different hardware platforms including CPU and a combination of CPU and GPU. Both NVIDIA and AMD GPUs have been used for performance evaluation. GPUs provide significant speedup over the parallel CPU version.

【 授权许可】

   
2014 Leeser et al.; licensee BioMed Central Ltd.

【 图 表 】

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

【 参考文献 】

• [1]Zhao X, Hu J-j, Zhang P: Gpu-based 3d cone beam ct image reconstruction for large data volume. Int J Biomed Imaging 2009, 2009:149079.
• [2]Feldkamp LA, Davis LC, Kress JW: Practical cone-beam algorithm. J Opt Soc Am 1984, 1:612-619.
• [3]Mueller K, Xu F: Practical consideration for gpu-accelerated ct. IEEE Int Symp Biomed Imaging 2006, 11:1184.
• [4]Fessler J: Image reconstruction toolbox. [http://web.eecs.umich.edu/~fessler/irt/fessler.tgz webcite]
• [5]Cbct open source software [http://www.coe.neu.edu/Research/rcl//projects/CBCT.php webcite]
• [6]Ino F, Yoshida S, Hagihara K: RGBA packing for fast cone beam reconstruction on the GPU. In SPIE Medical Imaging. International Society for Optics and Photonics; 2009:725858-725858.
• [7]de O Sandes EF, de Melo AC: Retrieving smith-waterman alignments with optimizations for megabase biological sequences using GPU. Parallel Distributed Syst IEEE Trans 2013, 24(5):1009-1021.
• [8]Sukhwani B, Herbordt MC: Gpu acceleration of a production molecular docking code. In Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units. New York: ACM; 2009:19-27.
• [9]Liu C-M, Wong T, Wu E, Luo R, Yiu S-M, Li Y, Wang B, Yu C, Chu X, Zhao K, Li R, Lam TW: SOAP3: ultra-fast gpu-based parallel alignment tool for short reads. Bioinformatics 2012, 28(6):878-879.
• [10]Ayres DL, Darling A, Zwickl DJ, Beerli P, Holder MT, Lewis PO, Huelsenbeck JP, Ronquist F, Swofford DL, Cummings MP, Rambaut A, Suchard MA: Beagle: An application programming interface and high-performance computing library for statistical phylogenetics. System Biol 2012, 61(1):170.
• [11]Jie L, Li K, Shi L, Liu R, Mei J: Accelerating solidification process simulation for large-sized system of liquid metal atoms using GPU with CUDA. J Comput Phys 2014, 257 Part A(0):521-535.
• [12]Valim N, Brock J, Leeser M, Niedre M: The effect of temporal impulse response on experimental reduction of photon scatter in time-resolved diffuse optical tomography. Phys Med Biol 2013, 58(2):335.
• [13]Okuyama T, Okita M, Abe T, Asai Y, Kitano H, Nomura T, Hagihara K: Accelerating ODE-based simulation of general and heterogeneous biophysical models using a GPU. Parallel Distributed Syst EEE Trans 2014, 25(8):1966-1975.
• [14]Leeser M, Ramachandran J, Wahl T, Yablonski D: OpenCL floating point software on heterogeneous architectures–portable or not. Workshop on Numerical Software Verification (NSV) 2012. Available from [http://www.ccs.neu.edu/home/wahl/Research/FPA-Heterogeneous/ webcite]
• [15]Xiao S, Bresler Y, Munson Jr DC: Fast Feldkamp algorithm for cone-beam computer tomography. In Image Processing, International Conference on (ICIP), Volume 2. New York: IEEE; 2003:819-819.
• [16]Basu S, Bresler Y: O(n2log2n) filtered backprojection reconstruction algorithm for tomography. IEEE Trans Image Process 2000, 9:10.
• [17]Rodet T, Noo F, Defrise M: The cone-beam algorithm of feldkamp, davis, and kress preserves oblique line integrals. Med Phys 2004, 31:1972.
• [18]Cabral B, Cam N, Foran J: Accelerated volume rendering and tomographic reconstruction using texture mapping hardware. In Proceedings of the 1994 Symposium on Volume Visualization. New York: ACM; 1994:91-98.
• [19]Mueller K, Xu F: Real-time 3d computed tomographic reconstruction using commodity graphics hardware. Phys Med Biol 2007, 52:3405.
• [20]Noël PB, Walczak AM, Xu J, Corso JJ, Hoffmann KR, Schafer S: GPU-based cone beam computed tomography. Comput Methods Prog Biomed 2010, 98(3):271-277.
• [21]Knaup M, Steckmann S, Kachelriess M: GPU-based parallel-beam and cone-beam forward-and backprojection using CUDA. In Nuclear Science Symposium Conference Record (NSS). New York: IEEE; 2008:5153-5157.
• [22]Mueller K, Xu F, Neophytou N: Why do commodity graphics hardware boards (GPUs) work so well for acceleration of computed tomography? In Electronic Imaging 2007. International Society for Optics and Photonics; 2007:64980-64980.
• [23]Jomier J, Rit S, Oliva MV: Rtk: The reconstruction toolkit. [http://www.kitware.com/source/home/post/115 webcite]
• [24]National Library of Medicine: Insight Segmentation and Registration Toolkit (ITK). [http://www.itk.org/ webcite]
• [25]Noël PB, Walczak A, Hoffmann KR, Xu J, Corso JJ, Schafer S: Clinical evaluation of gpu-based cone beam computed tomography. Proc. High-Performance Comput Biomed Image Anal 2008. [http://www.miccai.org/ webcite]
• [26]Mukherjee S, Moore N, Brock J, Leeser M: CUDA and OpenCL implementations of 3D CT reconstruction for biomedical imaging. In High Performance Extreme Computing (HPEC), Conference On. New York: IEEE; 2012:1-6.
• [27]OpenMP: OpenMP Standard Version 3.1. [http://www.openmp.org/mp-documents/OpenMP3.1.pdf webcite]
• [28]NVIDIA: CUDA CUFFT Library. [http://docs.nvidia.com/cuda/cufft/index.html webcite]