Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

MACHINE LEARNING

Computational staining of unlabelled tissue

Nature Biomedical Engineering volume 3, pages 425–426 (2019)Cite this article

Subjects

Histologically stained slides can be generated from unlabelled autofluorescence images of tissue samples via deep learning.

By comparing autofluorescence images of unstained tissue slides against ground-truth slides of the corresponding histopathology stains, Ozcan and co-authors’ deep-learning model (a deep neural network trained using a generative adversarial approach) statistically ‘learns’ the histological staining patterns routinely performed by pathologists. The ‘learned’ computational stains were then applied to unstained tissue, and compared to experimental stains (same piece of tissue stained with its corresponding histochemical stain) by expert pathologists, who were blinded to the source of the stained tissue (computational or experimental; they saw only a digital representation of both stains). The computational stains performed comparably or better than the experimental stains in a small blinded review, where quality of nuclear detail, cytoplasmic detail, extracellular matrix and overall stain quality were assessed. However, as noted by the authors, to determine its clinical utility this computational-staining method needs to be validated in larger and more comprehensive human trials. In addition, they note that this technology could produce more histochemical computational stains by extending the autofluorescence of the tissues through additional excitation and emission filters (the authors have shown this for the detection of melanin1). Moreover, autofluorescence may also be used to identify other components of the tissue (for example, extracellular vesicles, via label-free nonlinear microscopy of fresh tissue2) that are not visualized via routine histochemical stains.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Rent or buy this article

Get just this article for as long as you need it

$39.95

Learn more

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Histological staining (top) and computational staining (bottom) of tissue slides.

References

  1. Rivenson, Y. et al. Nat. Biomed. Eng. https://doi.org/10.1038/s41551-019-0362-y (2019).

  2. Sun, Y. et al. Sci. Adv. 4, eaau5603 (2018).

    Article  Google Scholar 

  3. Fereidouni, F. et al. Nat. Biomed. Eng. 1, 957–966 (2017).

    Article  Google Scholar 

  4. Glaser, A. K. et al. Nat. Biomed. Eng. 1, 0084 (2017).

    Article  Google Scholar 

  5. Beck, A. H. et al. Sci. Transl. Med. 3, 108ra113 (2011).

    Article  Google Scholar 

  6. Venet, L., Pati, S., Yushkevich, P. & Bakas, S. Preprint at https://arxiv.org/abs/1904.11929 (2019).

  7. Zwanenburg, A., Leger, S., Vallières, M. & Löck, S. Preprint at https://arxiv.org/abs/1612.07003 (2016).

  8. Binder, Z. A. et al. Cancer Cell 34, 163–177 (2018).

    Article  CAS  Google Scholar 

  9. Bakas, S. et al. Clin. Cancer Res. 23, 4724–4734 (2017).

    Article  CAS  Google Scholar 

  10. Akbari, H. et al. Neuro Oncol. 20, 1068–1079 (2018).

    Article  Google Scholar 

  11. Doyle, S., Feldman, M., Tomaszewski, J. & Madabhushi, A. IEEE Trans. Biomed. Eng. 59, 1205–1218 (2012).

    Article  Google Scholar 

  12. Whitney, J. et al. BMC Cancer 18, 610 (2018).

    Article  Google Scholar 

  13. Madabhushi, A. & Lee, G. Med. Image Anal. 33, 170–175 (2016).

    Article  Google Scholar 

  14. Goodfellow, I., Bengio, Y. & Courville, A. (eds) Deep Learning (MIT Press, 2016).

Download references

Author information

Authors and Affiliations

  1. Center for Biomedical Image Computing and Analytics, Richards Medical Research Laboratories, University of Pennsylvania, Philadelphia, PA, USA

    Spyridon Bakas & Michael D. Feldman

  2. Department of Pathology and Laboratory Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

    Spyridon Bakas & Michael D. Feldman

  3. Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

    Spyridon Bakas

Authors
  1. Spyridon Bakas
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Michael D. Feldman
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding authors

Correspondence to Spyridon Bakas or Michael D. Feldman.

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bakas, S., Feldman, M.D. Computational staining of unlabelled tissue. Nat Biomed Eng 3, 425–426 (2019). https://doi.org/10.1038/s41551-019-0414-3

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41551-019-0414-3

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing