Abstract
FlowDenoising is a new approach to noise reduction in biological volumes obtained with three-dimensional electron microscopy (3DEM). Its abilities to enhance the structural features stem from the fact that an anisotropic Gaussian filtering is steered according to the local structures. To this end, the Optical Flow (OF) among consecutive slices is estimated, which is the most computationally expensive step in this approach. In this article, a hybrid CPU/GPU implementation of FlowDenoising is introduced and evaluated. It exploits parallel computing by distributing the workload among multiple cores and takes advantage of the massive processing in GPUs to accelerate the OF estimation. The hybrid implementation provides remarkable speed-up factors and an important reduction of the processing time, which is particularly relevant for the denoising of huge volumes typically found in 3DEM.
Similar content being viewed by others
Data availability
The data of this study will be available from the corresponding authors on reasonable request. Code will be available through the github of the authors (https://github.com/microscopy-processing/FlowDenoising).
Notes
Available at https://github.com/microscopy-processing/FlowDenoising.
Notice that we have used the NumPy indexing notation.
References
Flynn M (1972) Some computer organizations and their effectiveness. IEEE Trans Comput C–21:948–960
Hennessy JL, Patterson DA (2019) Computer architecture. A quantitative approach. 6th Ed. Morgan Kauffman, USA
Brooks D (2011) CPUs, GPUs, and hybrid computing. IEEE Micro 31:4–6
Agulleiro JI, Vazquez F, Garzón EM, Fernandez JJ (2012) Hybrid computing: CPU+GPU co-processing and its application to tomographic reconstruction. Ultramicroscopy 115:109–114. https://doi.org/10.1016/j.ultramic.2012.02.003
Eisenstein M (2023) Seven technologies to watch in 2023. Nature 613:794–797
Peddie CJ et al (2022) Volume electron microscopy. Nat Rev Methods Prim 2:51. https://doi.org/10.1038/s43586-022-00131-9
Fernandez JJ, Martinez-Sanchez A (2022) Computational methods for three-dimensional electron microscopy (3DEM). Comput Methods Programs Biomed 225:107039. https://doi.org/10.1016/j.cmpb.2022.107039
Fernandez JJ (2012) Computational methods for electron tomography. Micron 43:1010–1030. https://doi.org/10.1016/j.micron.2012.05.003
Fernandez JJ, Li S (2003) An improved algorithm for anisotropic nonlinear diffusion for denoising cryo-tomograms. J Struct Biol 144:152–161. https://doi.org/10.1016/j.jsb.2003.09.010
Fernandez JJ, Li S (2005) Anisotropic nonlinear filtering of cellular structures in cryo-electron tomography. Comput Sci Eng 7(5):54–61. https://doi.org/10.1109/MCSE.2005.89
Moreno JJ, Martinez-Sanchez A, Martinez JA, Garzon EM, Fernandez JJ (2018) TomoEED: fast edge-enhancing denoising of tomographic volumes. Bioinformatics 34:3776–3778. https://doi.org/10.1093/bioinformatics/bty435
Lehtinen J, Munkberg J, Hasselgren J, Laine S, Karras T, Aittala M, Aila T (2018) Noise2Noise: learning image restoration without clean data. In: Dy J, Krause A (eds) Proceedings of the 35th International Conference on Machine Learning, vol. 80, pp 2965–2974
Krull A, Buchholz T-O, Jug F (2019) Noise2void-learning denoising from single noisy images. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 2129–2137. https://doi.org/10.1109/CVPR.2019.00223
Buchholz T-O, Krull A, Shahidi R, Pigino G, Jekely G, Jug F (2019) Content-aware image restoration for electron microscopy. In: Muller-Reichert T, Pigino G (eds) Three-dimensional electron microscopy. Methods in cell biology, vol 152. Academic Press, Cambridge, pp 277–289
Tabik S, Garzón EM, García I, Fernandez JJ (2007) High performance noise reduction for biomedical multidimensional data. Digit Signal Process 17:724–736. https://doi.org/10.1016/j.dsp.2006.11.004
Fernandez JJ (2008) High performance computing in structural determination by electron cryomicroscopy. J Struct Biol 164:1–6. https://doi.org/10.1016/j.jsb.2008.07.005
Cuomo S, Michele PD, Piccialli F (2014) 3D data denoising via Nonlocal means filter by using parallel GPU strategies. Comput Math Methods Med 164:523862
Tabik S, Murarasu A, Romero LF (2014) Anisotropic nonlinear diffusion for filtering 3d images on gpus. In: 2014 IEEE International Conference on Cluster Computing (CLUSTER). https://doi.org/10.1109/CLUSTER.2014.6968786
Kwon K, Kim MS, Shin BS (2016) A fast 3D adaptive bilateral filter for ultrasound volume visualization. Comput Methods Programs Biomed 133:25–34. https://doi.org/10.1016/j.cmpb.2016.05.008
Yano K, Sugimoto K, Kamata S-i (2018) GPU-friendly Approximate Bilateral Filter for 3D Volume Data. In: 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC), pp 2054–2058. https://doi.org/10.23919/APSIPA.2018.8659773
Steyer AM, Ruhwedel T, Nardis C, Werner HB, Nave KA, Möbius W (2020) Pathology of myelinated axons in the PLP-deficient mouse model of spastic paraplegia type 2 revealed by volume imaging using focused ion beam-scanning electron microscopy. J Struct Biol 210:107492. https://doi.org/10.1016/j.jsb.2020.107492
Hennies J, Lleti JMS, Schieber NL, Templin RM, Steyer AM, Schwab Y (2020) AMST: alignment to median smoothed template for focused ion beam scanning electron microscopy image stacks. Sci Rep 10:2004. https://doi.org/10.1038/s41598-020-58736-7
Uwizeye C, Decelle J, Jouneau P-H, Flori S, Gallet B, Keck J-B, Bo DD, Moriscot C, Seydoux C, Chevalier F, Schieber NL, Templin R, Allorent G, Courtois F, Curien G, Schwab Y, Schoehn G, Zeeman SC, Falconet D, Finazzi G (2021) Morphological bases of phytoplankton energy management and physiological responses unveiled by 3D subcellular imaging. Nat Commun 12:1049. https://doi.org/10.1038/s41467-021-21314-0
González-Ruiz V, Fernández-Fernández MR, Fernandez JJ (2023) Structure-preserving gaussian denoising of FIB-SEM volumes. Ultramicroscopy 246:113674. https://doi.org/10.1016/j.ultramic.2022.113674
González-Ruiz V, Fernandez J-J (2023) Flowdenoising: structure-preserving denoising in 3d electron microscopy (3dem). SoftwareX 23:101413. https://doi.org/10.1016/j.softx.2023.101413
Farnebäck G (2003) Two-frame motion estimation based on polynomial expansion. In: Scandinavian Conference on Image Analysis, pp 363–370. https://doi.org/10.1007/3-540-45103-X_50
Bykov YS et al (2017) The structure of the COPI coat determined within the cell. eLife 6:32493. https://doi.org/10.7554/eLife.32493
Knott G, Rosset S, Cantoni M (2011) Focussed ion beam milling and scanning electron microscopy of brain tissue. JoVE (Journal of Visualized Experiments) 53:2588. https://doi.org/10.3791/2588
Lucchi A, Li Y, Fua P (2013) Learning for structured prediction using approximate subgradient descent with working sets. In: 2013 IEEE Conference on Computer Vision and Pattern Recognition, pp 1987–1994. https://doi.org/10.1109/CVPR.2013.259
Wietrzynski W et al (2020) Charting the native architecture of chlamydomonas thylakoid membranes with single-molecule precision. eLife 9:53740. https://doi.org/10.7554/eLife.53740
Acknowledgements
Work supported by MCIN/AEI/10.13039/501100011033, “ERDF A way of making Europe” and by the “European Union NextGenerationEU/PRTR” through grants PID2021-123278OB-I00, TED2021-132020B-I00, PID2022-139071NB-I00 and PDC2022-133370-I00.
Funding
This work was supported by Spanish Ministry of Science and Innovation MCIN/AEI/10.13039/501100011033, “ERDF A way of making Europe” and by the “European Union NextGenerationEU/PRTR” through grants PID2021-123278OB-I00, TED2021-132020B-I00, PID2022-139071NB-I00 and PDC2022-133370-I00.
Author information
Authors and Affiliations
Contributions
VGR and JJF contributed to Conceptualization, Investigation, Software, Validation, Writing, and Funding Acquisition. JJM contributed to Investigation and Software.
Corresponding authors
Ethics declarations
Conflict of interest
The authors have no competing interests to declare that are relevant to the content of this article.
Ethical approval
Not applicable.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
González-Ruiz, V., Moreno, J.J. & Fernández, J.J. Acceleration of 3D feature-enhancing noise filtering in hybrid CPU/GPU systems. J Supercomput (2024). https://doi.org/10.1007/s11227-024-05928-x
Accepted:
Published:
DOI: https://doi.org/10.1007/s11227-024-05928-x