The DOI matches its Crossref record; 19 of the 19 references it cites were found in a registry or an index.
A distributed multi-GPU system for high speed electron microscopic tomographic reconstruction
Open accessAbstract
Full resolution electron microscopic tomographic (EMT) reconstruction of large-scale tilt series requires significant computing power. The desire to perform multiple cycles of iterative reconstruction and realignment dramatically increases the pressing need to improve reconstruction performance. This has motivated us to develop a distributed multi-GPU (graphics processing unit) system to provide the required computing power for rapid constrained, iterative reconstructions of very large three-dimensional (3D) volumes. The participating GPUs reconstruct segments of the volume in parallel, and subsequently, the segments are assembled to form the complete 3D volume. Owing to its power and versatility, the CUDA (NVIDIA, USA) platform was selected for GPU implementation of the EMT reconstruction. For a system containing 10 GPUs provided by 5 GTX295 cards, 10 cycles of SIRT reconstruction for a tomogram of 4096(2) × 512 voxels from an input tilt series containing 122 projection images of 4096(2) pixels (single precision float) takes a total of 1845 s of which 1032 s are for computation with the remainder being the system overhead. The same system takes only 39 s total to reconstruct 1024(2) × 256 voxels from 122 1024(2) pixel projections. While the system overhead is non-trivial, performance analysis indicates that adding extra GPUs to the system would lead to steadily enhanced overall performance. Therefore, this system can be easily expanded to generate superior computing power for very large tomographic reconstructions and especially to empower iterative cycles of reconstruction and realignment.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Noel ve diğerlerinin çalışmasında konik ışınlı bilgisayarlı tomografi ile 3B görüntüler elde edilmektedir.
“Konik ışınlı (cone beam) bilgisayarlı (computed) tomografi ile 3B görüntüler elde edilmektedir.”
From GPU Programlamada CUDA Platformu Kullanılan Paralel Görüntü İşleme Çalışmalarının İncelenmesi (2020) · cites this paper as [23]
The paper does reconstruct 3D volumes (Section 5), but it describes the method as electron microscopic tomography using a transmission electron microscope. Section 3 specifies a parallel-beam projection assumption, contrary to the attribution of cone-beam computed tomography. Page numbers were not available in our copy. (LLM, conf=0.84)
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredNoel ve diğerleri [23] 2010 yılında ağırlıklandırılmış filtrelenmiş geriye izdüşüm ile görüntü geriçatma işlemini CUDA kullanarak çift hassasiyetli kayan noktalı mimari ve basit mimari ile uygulamışlardır.
“Çalışmada ağırlıklandırılmış filtrelenmiş geriye izdüşüm ile görüntü geriçatma işlemi CUDA kullanarak çift hassasiyetli kayan noktalı mimari ve basit mimari ile uygulanmaktadır.”
From GPU Programlamada CUDA Platformu Kullanılan Paralel Görüntü İşleme Çalışmalarının İncelenmesi (2020) · cites this paper as [23]
The Introduction and §4 describe CUDA-based tomographic reconstruction, supporting that part of the claim. But §2 identifies the paper’s own methods as iterative SIRT and SART, not weighted filtered back-projection; §5 says its input and output data are single-precision, not double-precision. The paper also describes a distributed multi-GPU system rather than a simple architecture. (LLM, conf=0.91)
Experimental — accuracy not yet measuredİncelenen görüntü geriçatma çalışmalarında ileri ve geri izdüşüm yöntemi CUDA ile paralelleştirmede en çok kullanılan yöntemdir.
“İncelenen çalışmalarda görüntü geriçatmada ileri ve geri izdüşüm yönteminin CUDA ile paralelleştirmede en çok kullanılan yöntem olduğu görülmektedir [22, 23, 24, 25].”
From GPU Programlamada CUDA Platformu Kullanılan Paralel Görüntü İşleme Çalışmalarının İncelenmesi (2020) · cites this paper as [23]
Sections 3 and 4 (page numbers unavailable) describe forward- and back-projection as parallelized computations in this paper’s CUDA reconstruction system. The paper supports this as an example of the approach, but does not establish that it is the most commonly used method across the studies examined by the citing authors. (LLM, conf=0.85)
Experimental — accuracy not yet measuredNoel ve diğerlerinin çalışmasında çift hassasiyetli kayan noktalı mimaride, basit mimari ile karşılaştırıldığında %60 oranında hızlanma görülmüştür.
“Çift hassasiyetli kayan noktalı mimaride, basit mimari ile karşılaştırıldığında %60 oranında hızlanma görülmüştür.”
From GPU Programlamada CUDA Platformu Kullanılan Paralel Görüntü İşleme Çalışmalarının İncelenmesi (2020) · cites this paper as [23]
The System verification and performance tests section reports a nearly seven-fold speedup when the system is expanded to 10 GPUs, but does not compare double-precision and simple architectures. Searches for double precision, single precision, and a 60% speedup found no passage supporting the claimed comparison. (LLM, conf=0.76)
Experimental — accuracy not yet measuredNoel ve diğerlerinin çalışmasında standart yöntemlerle 256x256x256 boyutlarındaki bir görüntünün geriçatımı 25 dakikanın üzerinde zaman almaktadır.
“Standart yöntemlerle 256x256x256 boyutlarındaki bir görüntünün geriçatımı 25 dakikanın üzerinde zaman almaktadır.”
From GPU Programlamada CUDA Platformu Kullanılan Paralel Görüntü İşleme Çalışmalarının İncelenmesi (2020) · cites this paper as [23]
The performance tests (Section 5; page unavailable) report reconstruction times for GPU systems and different volume sizes, not a conventional-method reconstruction of a 256 × 256 × 256 volume. Searches for that size and a time exceeding 25 minutes found no passage establishing the claimed comparison. (LLM, conf=0.77)
Experimental — accuracy not yet measured
Noel ve diğerlerinin çalışmasında işlem zamanı 256x256x256 görüntülerde 3.2 saniyeye, 512x512x512 görüntülerde 8.5 saniyeye düşürülmüştür.
“İşlem zamanı 256x256x256 görüntülerde 3.2 saniyeye, 512x512x512 görüntülerde 8.5 saniyeye düşürülmüştür.”
From GPU Programlamada CUDA Platformu Kullanılan Paralel Görüntü İşleme Çalışmalarının İncelenmesi (2020) · cites this paper as [23]
Section 5 (page unavailable) reports reconstruction times for larger volumes, including less than 40 seconds for 1024² × 256 voxels and less than six minutes for 2048² × 512 voxels. Searches for 3.2 and 8.5 seconds and for the claimed 256³ and 512³ image sizes found no matching performance results in our copy. (LLM, conf=0.76)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Andersen, A. H., & Kak, A. C. (1984) · Ultrasonic Imaging · doi:10.1177/016173468400600107 · Cited by 1,502
Iterative methods for the three-dimensional reconstruction of an object from projections
Gilbert, P. F. C. (1972) · Journal of Theoretical Biology · doi:10.1016/0022-5193(72)90180-4 · Cited by 1,427
Andersen, A. H. (1984) · Ultrasonic Imaging · doi:10.1016/0161-7346(84)90008-7 · Cited by 1,162
Zheng, S. Q., Keszthelyi, B., Branlund, E., Lyle, J. M., Braunfeld, M. B., Sedat, J. W., & Agard, D. A. (2006) · Journal of Structural Biology · doi:10.1016/j.jsb.2006.06.005 · Cited by 206
Self-citationWinkler, H., & Taylor, K. A. (2005) · Ultramicroscopy · doi:10.1016/j.ultramic.2005.07.007 · Cited by 203
Fully automated, sequential tilt-series acquisition with Leginon
Suloway, C. J. M., Shi, J. R., Cheng, A., Pulokas, J., Carragher, B. O., Potter, C. S., Zheng, S. Q., Agard, D. A., & Jensen, G. J. (2009) · Journal of Structural Biology · doi:10.1016/j.jsb.2009.03.019 · Cited by 182
Self-citationPerformance evaluation of image processing algorithms on the GPU
Castaño‐Díez, D., Moser, D., Schönegger, A., Pruggnaller, S., & Frangakis, A. S. (2008) · Journal of Structural Biology · doi:10.1016/j.jsb.2008.07.006 · Cited by 107
High-performance electron tomography of complex biological specimens
Fernández, J., Lawrence, A. F., Roca, J., García, I., Ellisman, M. H., & Carazo, J. M. (2002) · Journal of Structural Biology · doi:10.1016/s1047-8477(02)00017-5 · Cited by 88
How electron cryotomography is opening a new window onto prokaryotic ultrastructure
Jensen, G. J., & Briegel, A. (2007) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2007.03.002 · Cited by 82
Xu, W., Xu, F., Jones, M., Keszthelyi, B., Sedat, J., Agard, D., & Mueller, K. (2010) · Journal of Structural Biology · doi:10.1016/j.jsb.2010.03.018 · Cited by 71
Self-citationA novel dual-axis iterative algorithm for electron tomography
Tong, J. R., Arslan, I., & Midgley, P. A. (2005) · Journal of Structural Biology · doi:10.1016/j.jsb.2005.10.005 · Cited by 68
Implementation and performance evaluation of reconstruction algorithms on graphics processors
Castaño Díez, D., Mueller, H., & Frangakis, A. S. (2007) · Journal of Structural Biology · doi:10.1016/j.jsb.2006.08.010 · Cited by 68
Marabini, R., Rietzel, E., Schroeder, R. R., et al. (1997) · Journal of Structural Biology · doi:10.1006/jsbi.1997.3923 · Cited by 65
On the efficiency of iterative ordered subset reconstruction algorithms for acceleration on GPUs
Xu, F., Xu, W., Jones, M., Keszthelyi, B., Sedat, J., Agard, D., & Mueller, K. (2010) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2009.09.003 · Cited by 63
Self-citationMulti GPU implementation of iterative tomographic reconstruction algorithms
Jang, B., Kaeli, D. R., Do, S., et al. (2009) · IEEE International Symposium on Biomedical Imaging: From Nano to Macro (ISBI) · doi:10.1109/isbi.2009.5193014 · Cited by 57
A matrix approach to tomographic reconstruction and its implementation on GPUs
Vázquez, F., Garzón, E. M., & Fernández, J. J. (2010) · Journal of Structural Biology · doi:10.1016/j.jsb.2010.01.021 · Cited by 39
Efficient parallel implementation of iterative reconstruction algorithms for electron tomography
Fernández, J., Gordon, D., & Gordon, R. (2007) · Journal of Parallel and Distributed Computing · doi:10.1016/j.jpdc.2007.09.003 · Cited by 25
Vectorization with SIMD extensions speeds up reconstruction in electron tomography
Agulleiro, J. I., Garzón, E. M., ́a, I. G., & Fernández, J. (2010) · Journal of Structural Biology · doi:10.1016/j.jsb.2010.01.008 · Cited by 24
Dual-axis target mapping and automated sequential acquisition of dual-axis EM tomographic data
Zheng, S. Q., Matsuda, A., Braunfeld, M. B., et al. (2009) · Journal of Structural Biology · doi:10.1016/j.jsb.2009.06.010 · Cited by 13
Self-citation
Flagged first, then by citations
Research works printed but not linked: 2
- B. Jang, D. Kaeli, S. Do, H. Pien, in: Proceedings of the Sixth IEEE International Conference on Symposium on Biomedical Imaging, 2009, pp. 185.
- Chen, 1992, Proc. SPIE Int. Soc. Opt. Eng., 1660, 784
Cited by
Sorted by citations
Hybrid computing: CPU+GPU co-processing and its application to tomographic reconstruction
Agulleiro, J. I., Vázquez, F., Garzón, E. M., & Fernández, J. J. (2012) · Ultramicroscopy · doi:10.1016/j.ultramic.2012.02.003 · Cited by 34
GPU Programlamada CUDA Platformu Kullanılan Paralel Görüntü İşleme Çalışmalarının İncelenmesi
AYDIN, S., SAMET, R., & BAY, Ö. F. (2020) · Politeknik Dergisi · doi:10.2339/politeknik.563767 · Cited by 3
“Zheng ve diğerleri [24] çalışmalarında iteratif geriçatma yöntemini GPU sayısını artırarak daha kısa sürede gerçekleştirmektedirler.”
Metadata
- Authors
- Zheng, S. Q., Branlund, E., Kesthelyi, B., Braunfeld, M. B., Cheng, Y., Sedat, J. W., & Agard, D. A.
- Affiliations
- Howard Hughes Medical Institute (US)
- Corresponding author
- Agard, David A.
- Venue
- Ultramicroscopy
- Publisher
- Elsevier BV
- Published
- 2011-07
- Volume
- 111
- Issue
- 8
- Pages
- 1137-1143
- ISSN
- 0304-3991 (ISSN-L)
- PMC
- PMC3190582
- OpenAlex
- W2096397551
- Cited by
- 13
- References
- 19
- Citations by year
- 2019: 1 · 2018: 2 · 2017: 1 · 2014: 2 · 2013: 2 · 2012: 5
- FWCI
- 1.87
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Locations
- Ultramicroscopy (publishedVersion)
- PubMed (publishedVersion)
- Europe PMC (PubMed Central) (acceptedVersion)
- PubMed Central (acceptedVersion)
- Funders
- NIH (1S10RR026814-01)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Structural Biology › Advanced Electron Microscopy Techniques and Applications
- Physical Sciences › Computer Science › Hardware and Architecture › Parallel Computing and Optimization Techniques
- Physical Sciences › Computer Science › Computer Graphics and Computer-Aided Design › Computer Graphics and Visualization Techniques
- Keywords
- electron tomography, tomographic reconstruction, multi-GPU computing, distributed computing, CUDA, iterative reconstruction, SIRT, tilt series, GPU parallel processing, volumetric reconstruction
- MeSH terms
- Algorithms, Animals, Computer Communication Networks, Computer Graphics, Drosophila, Image Processing, Computer-Assisted, Centrosome, Imaging, Three-Dimensional, Electron Microscope Tomography
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
