Trace: a high-throughput tomographic reconstruction engine for large-scale datasets
Abstract
Abstract Background Modern synchrotron light sources and detectors produce data at such scale and complexity that large-scale computation is required to unleash their full power. One of the widely used imaging techniques that generates data at tens of gigabytes per second is computed tomography (CT). Although CT experiments result in rapid data generation, the analysis and reconstruction of the collected data may require hours or even days of computation time with a medium-sized workstation, which hinders the scientific progress that relies on the results of analysis. Methods We present Trace, a data-intensive computing engine that we have developed to enable high-performance implementation of iterative tomographic reconstruction algorithms for parallel computers. Trace provides fine-grained reconstruction of tomography datasets using both (thread-level) shared memory and (process-level) distributed memory parallelization. Trace utilizes a special data structure called replicated reconstruction object to maximize application performance. We also present the optimizations that we apply to the replicated reconstruction objects and evaluate them using tomography datasets collected at the Advanced Photon Source. Results Our experimental evaluations show that our optimizations and parallelization techniques can provide 158× speedup using 32 compute nodes (384 cores) over a single-core configuration and decrease the end-to-end processing time of a largemore »
- Authors:
- Publication Date:
- Research Org.:
- Argonne National Laboratory (ANL), Argonne, IL (United States)
- Sponsoring Org.:
- USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)
- OSTI Identifier:
- 1344024
- Alternate Identifier(s):
- OSTI ID: 1375811
- Grant/Contract Number:
- AC02-06CH11357
- Resource Type:
- Published Article
- Journal Name:
- Advanced Structural and Chemical Imaging
- Additional Journal Information:
- Journal Name: Advanced Structural and Chemical Imaging Journal Volume: 3 Journal Issue: 1; Journal ID: ISSN 2198-0926
- Publisher:
- Springer Science + Business Media
- Country of Publication:
- Germany
- Language:
- English
- Subject:
- 97 MATHEMATICS AND COMPUTING; 46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND TECHNOLOGY; Big Data; High-Throughput; Reconstruction; Tomography
Citation Formats
Bicer, Tekin, Gürsoy, Doğa, Andrade, Vincent De, Kettimuthu, Rajkumar, Scullin, William, Carlo, Francesco De, and Foster, Ian T. Trace: a high-throughput tomographic reconstruction engine for large-scale datasets. Germany: N. p., 2017.
Web. doi:10.1186/s40679-017-0040-7.
Bicer, Tekin, Gürsoy, Doğa, Andrade, Vincent De, Kettimuthu, Rajkumar, Scullin, William, Carlo, Francesco De, & Foster, Ian T. Trace: a high-throughput tomographic reconstruction engine for large-scale datasets. Germany. https://doi.org/10.1186/s40679-017-0040-7
Bicer, Tekin, Gürsoy, Doğa, Andrade, Vincent De, Kettimuthu, Rajkumar, Scullin, William, Carlo, Francesco De, and Foster, Ian T. Sat .
"Trace: a high-throughput tomographic reconstruction engine for large-scale datasets". Germany. https://doi.org/10.1186/s40679-017-0040-7.
@article{osti_1344024,
title = {Trace: a high-throughput tomographic reconstruction engine for large-scale datasets},
author = {Bicer, Tekin and Gürsoy, Doğa and Andrade, Vincent De and Kettimuthu, Rajkumar and Scullin, William and Carlo, Francesco De and Foster, Ian T.},
abstractNote = {Abstract Background Modern synchrotron light sources and detectors produce data at such scale and complexity that large-scale computation is required to unleash their full power. One of the widely used imaging techniques that generates data at tens of gigabytes per second is computed tomography (CT). Although CT experiments result in rapid data generation, the analysis and reconstruction of the collected data may require hours or even days of computation time with a medium-sized workstation, which hinders the scientific progress that relies on the results of analysis. Methods We present Trace, a data-intensive computing engine that we have developed to enable high-performance implementation of iterative tomographic reconstruction algorithms for parallel computers. Trace provides fine-grained reconstruction of tomography datasets using both (thread-level) shared memory and (process-level) distributed memory parallelization. Trace utilizes a special data structure called replicated reconstruction object to maximize application performance. We also present the optimizations that we apply to the replicated reconstruction objects and evaluate them using tomography datasets collected at the Advanced Photon Source. Results Our experimental evaluations show that our optimizations and parallelization techniques can provide 158× speedup using 32 compute nodes (384 cores) over a single-core configuration and decrease the end-to-end processing time of a large sinogram (with 4501 × 1 × 22,400 dimensions) from 12.5 h to <5 min per iteration. Conclusion The proposed tomographic reconstruction engine can efficiently process large-scale tomographic data using many compute nodes and minimize reconstruction times.},
doi = {10.1186/s40679-017-0040-7},
journal = {Advanced Structural and Chemical Imaging},
number = 1,
volume = 3,
place = {Germany},
year = {Sat Jan 28 00:00:00 EST 2017},
month = {Sat Jan 28 00:00:00 EST 2017}
}
https://doi.org/10.1186/s40679-017-0040-7
Works referenced in this record:
TIMBIR: A Method for Time-Space Reconstruction From Interlaced Views
journal, June 2015
- Aditya Mohan, K.; Venkatakrishnan, S. V.; Gibbs, John W.
- IEEE Transactions on Computational Imaging, Vol. 1, Issue 2
Hybrid MPI-OpenMP Programming for Parallel OSEM PET Reconstruction
journal, October 2006
- Jones, M. D.; Yao, R.; Bhole, C. P.
- IEEE Transactions on Nuclear Science, Vol. 53, Issue 5
Fast tomographic reconstruction on multicore computers
journal, December 2010
- Agulleiro, J. I.; Fernandez, J. J.
- Bioinformatics, Vol. 27, Issue 4
Pushing the limits for medical image reconstruction on recent standard multicore processors
journal, June 2012
- Treibig, Jan; Hager, Georg; Hofmann, Hannes G.
- The International Journal of High Performance Computing Applications, Vol. 27, Issue 2
A new workflow for x-ray fluorescence tomography: MAPStoTomoPy
conference, September 2015
- Hong, Young Pyo; Chen, Si; Jacobsen, Chris
- SPIE Optical Engineering + Applications, SPIE Proceedings
IDEAL: Images Across Domains, Experiments, Algorithms and Learning
journal, September 2016
- Ushizima, Daniela M.; Bale, Hrishikesh A.; Bethel, E. Wes
- JOM, Vol. 68, Issue 11
A Map-Reduce System with an Alternate API for Multi-core Environments
conference, May 2010
- Jiang, Wei; Ravi, Vignesh T.; Agrawal, Gagan
- 2010 10th IEEE/ACM International Conference on Cluster, Cloud and Grid Computing
A comparative study of X-ray tomographic microscopy on shales at different synchrotron facilities: ALS, APS and SLS
journal, November 2012
- Kanitpanyacharoen, Waruntorn; Parkinson, Dilworth Y.; De Carlo, Francesco
- Journal of Synchrotron Radiation, Vol. 20, Issue 1
Maximum a posteriori estimation of crystallographic phases in X-ray diffraction tomography
journal, June 2015
- Gürsoy, Doĝa; Biçer, Tekin; Almer, Jonathan D.
- Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 373, Issue 2043
Workflow Management for Real-Time Analysis of Lightsource Experiments
conference, November 2014
- Deslippe, Jack; Essiari, Abdelilah; Patton, Simon J.
- 2014 9th Workshop on Workflows in Support of Large-Scale Science (WORKS)
Electron tomography based on a total variation minimization reconstruction technique
journal, February 2012
- Goris, B.; Van den Broek, W.; Batenburg, K. J.
- Ultramicroscopy, Vol. 113
Optimization of tomographic reconstruction workflows on geographically distributed resources
journal, June 2016
- Bicer, Tekin; Gürsoy, Dogˇa; Kettimuthu, Rajkumar
- Journal of Synchrotron Radiation, Vol. 23, Issue 4
Regridding reconstruction algorithm for real-time tomographic imaging
journal, September 2012
- Marone, F.; Stampanoni, M.
- Journal of Synchrotron Radiation, Vol. 19, Issue 6
A fast forward projection using multithreads for multirays on GPUs in medical image reconstruction: Fast forward projection on GPUs
journal, June 2011
- Chou, Cheng-Ying; Chuo, Yi-Yen; Hung, Yukai
- Medical Physics, Vol. 38, Issue 7
Hyperspectral image reconstruction for x-ray fluorescence tomography
journal, January 2015
- Gürsoy, Doǧa; Biçer, Tekin; Lanzirotti, Antonio
- Optics Express, Vol. 23, Issue 7
CAMERA: The Center for Advanced Mathematics for Energy Research Applications
journal, March 2015
- Donatelli, J.; Haranczyk, M.; Hexemer, A.
- Synchrotron Radiation News, Vol. 28, Issue 2
Time-resolved X-ray Tomography of Gasoline Direct Injection Sprays
journal, September 2015
- Duke, Daniel J.; Swantek, Andrew B.; Sovis, Nicolas M.
- SAE International Journal of Engines, Vol. 9, Issue 1
High performance model based image reconstruction
conference, January 2016
- Wang, Xiao; Sabne, Amit; Kisner, Sherman
- Proceedings of the 21st ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming - PPoPP '16
Real-time data-intensive computing
conference, January 2016
- Parkinson, Dilworth Y.; Beattie, Keith; Chen, Xian
- PROCEEDINGS OF THE 12TH INTERNATIONAL CONFERENCE ON SYNCHROTRON RADIATION INSTRUMENTATION – SRI2015, AIP Conference Proceedings
A data-parallel algorithm for iterative tomographic image reconstruction
conference, January 1999
- Johnson, C. A.; Sofer, A.
- Proceedings. Frontiers '99. Seventh Symposium on the Frontiers of Massively Parallel Computation
Data Analysis WorkbeNch ( DAWN )
journal, April 2015
- Basham, Mark; Filik, Jacob; Wharmby, Michael T.
- Journal of Synchrotron Radiation, Vol. 22, Issue 3
Accelerating advanced MRI reconstructions on GPUs
journal, October 2008
- Stone, S. S.; Haldar, J. P.; Tsao, S. C.
- Journal of Parallel and Distributed Computing, Vol. 68, Issue 10
CUDA optimization strategies for compute- and memory-bound neuroimaging algorithms
journal, June 2012
- Lee, Daren; Dinov, Ivo; Dong, Bin
- Computer Methods and Programs in Biomedicine, Vol. 106, Issue 3
Shared memory parallelization of data mining algorithms: techniques, programming interface, and performance
journal, January 2005
- Jin, Ruoming; Yang, Ge; Agrawal, G.
- IEEE Transactions on Knowledge and Data Engineering, Vol. 17, Issue 1, p. 71-89
A Fast CT Reconstruction Scheme for a General Multi-Core PC
journal, January 2007
- Zeng, Kai; Bai, Erwei; Wang, Ge
- International Journal of Biomedical Imaging, Vol. 2007
Fast calculation of the exact radiological path for a three-dimensional CT array: Technical Reports: 3D CT array path calculation
journal, March 1985
- Siddon, Robert L.
- Medical Physics, Vol. 12, Issue 2
The ASTRA Toolbox: A platform for advanced algorithm development in electron tomography
journal, October 2015
- van Aarle, Wim; Palenstijn, Willem Jan; De Beenhouwer, Jan
- Ultramicroscopy, Vol. 157
Spade: Decentralized orchestration of data movement and warehousing for physics experiments
conference, May 2015
- Patton, Simon; Samak, Taghrid; Tull, Craig E.
- 2015 IFIP/IEEE International Symposium on Integrated Network Management (IM)
Integration of TomoPy and the ASTRA toolbox for advanced processing and reconstruction of tomographic synchrotron data
journal, April 2016
- Pelt, Daniël M.; Gürsoy, Dogˇa; Palenstijn, Willem Jan
- Journal of Synchrotron Radiation, Vol. 23, Issue 3