CoChat
Sign Up
Cited by 17

The DOI matches its Crossref record; 2 of the 2 references it cites were found in a registry or an index.

An AI-based approach for detecting cells and microbial byproducts in low volume scanning electron microscope images of biofilms

Abeyrathna, Dilanga, Ashaduzzaman, Md, Malshe, Milind, Kalimuthu, Jawaharraj, Gadhamshetty, Venkataramana, Chundi, Parvathi, Subramaniam, Mahadevan·2022·Frontiers in Microbiology
DOI 10.3389/fmicb.2022.996400PMC PMC9751328
Open this paper in CoChat — free account

Abstract

Microbially induced corrosion (MIC) of metal surfaces caused by biofilms has wide-ranging consequences. Analysis of biofilm images to understand the distribution of morphological components in images such as microbial cells, MIC byproducts, and metal surfaces non-occluded by cells can provide insights into assessing the performance of coatings and developing new strategies for corrosion prevention. We present an automated approach based on self-supervised deep learning methods to analyze Scanning Electron Microscope (SEM) images and detect cells and MIC byproducts. The proposed approach develops models that can successfully detect cells, MIC byproducts, and non-occluded surface areas in SEM images with a high degree of accuracy using a low volume of data while requiring minimal expert manual effort for annotating images. We develop deep learning network pipelines involving both contrastive (Barlow Twins) and non-contrastive (MoCoV2) self-learning methods and generate models to classify image patches containing three labels—cells, MIC byproducts, and non-occluded surface areas. Our experimental results based on a dataset containing seven grayscale SEM images show that both Barlow Twin and MoCoV2 models outperform the state-of-the-art supervised learning models achieving prediction accuracy increases of approximately 8 and 6%, respectively. The self-supervised pipelines achieved this superior performance by requiring experts to annotate only ~10% of the input data. We also conducted a qualitative assessment of the proposed approach using experts and validated the classification outputs generated by the self-supervised models. This is perhaps the first attempt toward the application of self-supervised learning to classify biofilm image components and our results show that self-supervised learning methods are highly effective for this task while minimizing the expert annotation effort.

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.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.3389/fmicb.2022.996400

Cites

2 of 2 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Flow chart of the classification task.
Figure 1 Flow chart of the classification task.CC BY · PubMed Central
Figure 2 Output patches after super-resolution method. Top row contain patches from original image patches. Bottom row contains patches after applying super-resolution.
Figure 2 Output patches after super-resolution method. Top row contain patches from original image patches. Bottom row contains patches after applying super-resolution.CC BY · PubMed Central
Figure 3 Sample SEM images (first row) with their respective annotations (second row), where Byproduct, Cells, and Surface indicates by the colors pink, blue, and green, respectively.
Figure 3 Sample SEM images (first row) with their respective annotations (second row), where Byproduct, Cells, and Surface indicates by the colors pink, blue, and green, respectively.CC BY · PubMed Central
Figure 4 Object masking example. (A) Original image patch, and its object masks of (B) Byproduct, (C) Cell, and (D) Surface.
Figure 4 Object masking example. (A) Original image patch, and its object masks of (B) Byproduct, (C) Cell, and (D) Surface.CC BY · PubMed Central
Figure 5 Self-supervised frameworks (A) MoCoV2, (B) Barlow Twins.
Figure 5 Self-supervised frameworks (A) MoCoV2, (B) Barlow Twins.CC BY · PubMed Central
Figure 6 Interface for qualitative feedback. Image in the first row, first column shows an image patch followed by three columns of CAMs showing the true for class Surface, true for class Cells, and true for class Byproduct. Checkboxes and textboxes in the bottom half of the figure are available to input experts' agreements and feedback.
Figure 6 Interface for qualitative feedback. Image in the first row, first column shows an image patch followed by three columns of CAMs showing the true for class Surface, true for class Cells, and true for class Byproduct. Checkboxes and textboxes in the bottom half of the figure are available to input experts' agreements and feedback.CC BY · PubMed Central
Figure 7 Linear evaluation accuracy comparison of MoCoV2 and Barlow Twins models with different configurations of batch sizes and patch sizes. The first row of the y-axis indicates the patch sizes, the second row indicates batch sizes, and the third row indicates the corresponding self-supervised framework.
Figure 7 Linear evaluation accuracy comparison of MoCoV2 and Barlow Twins models with different configurations of batch sizes and patch sizes. The first row of the y-axis indicates the patch sizes, the second row indicates batch sizes, and the third row indicates the corresponding self-supervised framework.CC BY · PubMed Central

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.

A few free checks a day · papers already verified open instantly · results are public

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?

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

Research works printed but not linked: 44

  1. Atha, 2018, Evaluation of deep learning approaches based on convolutional neural networks for corrosion detection, Struct. Health Monitor, 17, 1110, 10.1177/1475921717737051 · DOI
  2. Azizi, 2021, Big self-supervised models advance medical image classification, arXiv preprint arXiv:2101.05224, 10.1109/ICCV48922.2021.00346 · DOI · Paper page
  3. Balestriero, 2022, Contrastive and non-contrastive self-supervised learning recover global and local spectral embedding methods, arXiv preprint arXiv:2205.11508, 10.48550/arXiv.2205.11508 · DOI
  4. Barlow, 1961, Possible principles underlying the transformation of sensory messages, Sensory Commun, 1, 217, 10.7551/mitpress/9780262518420.003.0013 · DOI
  5. Bay, 2006, Surf: speeded up robust features,”, 404
  6. Bogachev, 2018, Fast and simple tool for the quantification of biofilm-embedded cells sub-populations from fluorescent microscopic images, PLoS ONE, 13, e0193267, 10.1371/journal.pone.0193267 · DOI
  7. Bommanapally, 2021, Self-supervised learning approach to detect corrosion products in biofilm images,”, 3555, 10.1109/BIBM52615.2021.9669343 · DOI
  8. Buetti-Dinh, 2019, Deep neural networks outperform human expert's capacity in characterizing bioleaching bacterial biofilm composition, Biotechnol. Rep, 22, e00321, 10.1016/j.btre.2019.e00321 · DOI
  9. Chakravarthy, 2022, Semantic image segmentation using scant pixel annotations, Mach. Learn. Knowledge Extract, 4, 621, 10.3390/make4030029 · DOI
  10. Chen, A simple framework for contrastive learning of visual representations,”, 1597
  11. Chen, Improved baselines with momentum contrastive learning, arXiv preprint arXiv:2003.04297, 10.48550/arXiv.2003.04297 · DOI
  12. Chen, 2020, Exploring simple siamese representation learning, arXiv preprint arXiv:2011.10566, 10.1109/CVPR46437.2021.01549 · DOI · Paper page
  13. Chilkoor, 2020, Hexagonal boron nitride for sulfur corrosion inhibition, ACS Nano, 14, 14809, 10.1021/acsnano.0c03625 · DOI
  14. Ciga, 2021, Overcoming the limitations of patch-based learning to detect cancer in whole slide images, Sci. Rep, 11, 1, 10.1038/s41598-021-88494-z · DOI
  15. Dedysh, 2015, Methylocapsa palsarum sp. nov., a methanotroph isolated from a subarctic discontinuous permafrost ecosystem, Int. J. Syst. Evol. Microbiol, 65, 3618, 10.1099/ijsem.0.000465 · DOI
  16. Dong, 2021, A survey on deep learning and its applications, Comput. Sci. Rev, 40, 100379, 10.1016/j.cosrev.2021.100379 · DOI
  17. Grill, 2020, Bootstrap your own latent: a new approach to self-supervised learning, arXiv preprint arXiv:2006.07733, 10.48550/arXiv.2006.07733 · DOI · Paper page
  18. Hartmann, 2021, Quantitative image analysis of microbial communities with biofilmq, Nature microbiology, 6, 151, 10.1038/s41564-020-00817-4 · DOI
  19. He, 2020, Momentum contrast for unsupervised visual representation learning,”, 9729
  20. He, 2016, Deep residual learning for image recognition,”, 770
  21. Hough, 1962, Method and means for recognizing complex patterns, US Patent 3069654
  22. Huang, 2019, Deep Learning: Fundamentals, Theory and Applications, 10.1007/978-3-030-06073-2 · DOI
  23. Jaiswal, 2020, A survey on contrastive self-supervised learning, Technologies, 9, 2, 10.3390/technologies9010002 · DOI · Paper page
  24. Jamaludin, 2017, Self-supervised learning for spinal MRIs,”, 294, 10.1007/978-3-319-67558-9_34 · DOI
  25. Kornblith, 2019, Do better imagenet models transfer better?”, 2661
  26. Lam, 2017, Automated detection of diabetic retinopathy using deep learning,”, 147
  27. Li, 2004, Lazy snapping, ACM Trans. Graph, 23, 303, 10.1145/1015706.1015719 · DOI
  28. Liang, 2021, SwinIR: image restoration using swin transformer,”, 1833
  29. Little, 2020, Microbially influenced corrosion–any progress?, Corrosion Sci, 170, 108641, 10.1016/j.corsci.2020.108641 · DOI
  30. Mian, 2006, Three-dimensional model-based object recognition and segmentation in cluttered scenes, IEEE Trans. Pattern Anal. Mach. Intell, 28, 1584, 10.1109/TPAMI.2006.213 · DOI
  31. Ragi, 2021, Artificial intelligence-driven image analysis of bacterial cells and biofilms, arXiv:2112.01577, 10.48550/597ARXIV.2112.01577 · DOI
  32. Rueden, 2017, Imagej2: Imagej for the next generation of scientific image data, BMC Bioinformatics, 18, 529, 10.1186/s12859-017-1934-z · DOI
  33. Soda, 2009, Aggregation of classifiers for staining pattern recognition in antinuclear autoantibodies analysis, IEEE Trans. Inform. Technol. Biomed, 13, 322, 10.1109/TITB.2008.2010855 · DOI
  34. Stoean, 2018, Deep learning for metal corrosion control: can convolutional neural networks measure inhibitor efficiency?”, 387, 10.1109/SYNASC.2018.00065 · DOI
  35. Susarla, 2021, Corrosion resistance of sulfur-selenium alloy coatings, Adv. Mater, 33, 2104467, 10.1002/adma.202104467 · DOI
  36. Tajbakhsh, 2019, Surrogate supervision for medical image analysis: effective deep learning from limited quantities of labeled data,”, 1251, 10.1109/ISBI.2019.8759553 · DOI
  37. Tian, 2021, Understanding self-supervised learning dynamics without contrastive pairs,”, 10268
  38. Tsai, 2022, Multi-scale patch-based representation learning for image anomaly detection and segmentation,”, 3992
  39. Wang, 2021, Real-ESRGAN: training real-world blind super-resolution with pure synthetic data,”, 1905
  40. Zaidi, 2022, A survey of modern deep learning based object detection models, Digit. Signal Process, 126, 103514, 10.1016/j.dsp.2022.103514 · DOI
  41. Zbontar, 2021, Barlow twins: self-supervised learning via redundancy reduction, arXiv preprint arXiv:2103.03230, 10.48550/ARXIV.2103.03230 · DOI
  42. Zhang, 2021, Designing a practical degradation model for deep blind image super-resolution,”, 4791
  43. Zhou, 2016, Learning deep features for discriminative localization,”, 10.1109/CVPR.2016.319 · DOI · Paper page
  44. Zhou, 2017, Multi-Label Learning, 10.1007/978-1-4899-7687-1_910 · DOI

Cited by

Sorted by citations

See all 17

Metadata

Authors
Abeyrathna, D., Ashaduzzaman, M., Malshe, M., Kalimuthu, J., Gadhamshetty, V., Chundi, P., & Subramaniam, M.
Venue
Frontiers in Microbiology
Publisher
Frontiers Media SA
Published
2022-12-01
Volume
13
ISSN
1664-302X
DOI
10.3389/fmicb.2022.996400
PMC
PMC9751328
Cited by
17
References
2
Licence
creativecommons.org/licenses/by/4.0/ (vor)

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