EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Evaluation of Swin Transformer and knowledge transfer for denoising of super-resolution structured illumination microscopy data

Zafran Hussain Shah iD, Marcel Müller iD, Wolfgang Hübner iD, Tung-Cheng Wang, Daniel Telman, Thomas Huser iD, Wolfram Schenck iD

GigaScience · 2024

DOI: 10.1093/gigascience/giad109 ↗

Kurz gefasst

Fachartikel in der Zeitschrift GigaScience, 2024. Verfasst von Zafran Hussain Shah und 6 weiteren Autor:innen. Beteiligte Einrichtung: Faculty of Engineering and Mathematics, Bielefeld University of Applied Sciences and Arts, 33619 Bielefeld, Germany. Hauptthemen laut MEDLINE: Image Processing, Computer-Assisted, Microscopy. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 58 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Horizon 2020, Deutsche Forschungsgemeinschaft.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Convolutional neural network (CNN)–based methods have shown excellent performance in denoising and reconstruction of super-resolved structured illumination microscopy (SR-SIM) data. Therefore, CNN-based architectures have been the focus of existing studies. However, Swin Transformer, an alternative and recently proposed deep learning–based image restoration architecture, has not been fully investigated for denoising SR-SIM images. Furthermore, it has not been fully explored how well transfer learning strategies work for denoising SR-SIM images with different noise characteristics and recorded cell structures for these different types of deep learning–based methods. Currently, the scarcity of publicly available SR-SIM datasets limits the exploration of the performance and generalization capabilities of deep learning methods.

Results

In this work, we present SwinT-fairSIM, a novel method based on the Swin Transformer for restoring SR-SIM images with a low signal-to-noise ratio. The experimental results show that SwinT-fairSIM outperforms previous CNN-based denoising methods. Furthermore, as a second contribution, two types of transfer learning—namely, direct transfer and fine-tuning—were benchmarked in combination with SwinT-fairSIM and CNN-based methods for denoising SR-SIM data. Direct transfer did not prove to be a viable strategy, but fine-tuning produced results comparable to conventional training from scratch while saving computational time and potentially reducing the amount of training data required. As a third contribution, we publish four datasets of raw SIM images and already reconstructed SR-SIM images. These datasets cover two different types of cell structures, tubulin filaments and vesicle structures. Different noise levels are available for the tubulin filaments.

Conclusion

The SwinT-fairSIM method is well suited for denoising SR-SIM images. By fine-tuning, already trained models can be easily adapted to different noise characteristics and cell structures. Furthermore, the provided datasets are structured in a way that the research community can readily use them for research on denoising, super-resolution, and transfer learning strategies.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Image Processing, Computer-AssistedMicroscopyLightingTubulinNeural Networks, Computer

Stichwörter der Autor:innen

Convolutional neural networksDeep learningDenoisingFine-tuningFluorescence microscopyImage restorationStructured illumination microscopySwin TransformerSwinIRTransfer learningTransformer

Substanzen

Tubulin

Autor:innen und Institutionen

  • Faculty of Engineering and Mathematics, Bielefeld University of Applied Sciences and Arts, 33619 Bielefeld, Germany.
  • Faculty of Physics, Bielefeld University, 33615 Bielefeld, Germany.
  • Leica Microsystems CMS GmbH, 68165 Mannheim, Germany.

ORCID-Profile: Zafran Hussain Shah, Marcel Müller, Wolfgang Hübner, Thomas Huser, Wolfram Schenck

Förderung

  • Horizon 2020
  • Deutsche Forschungsgemeinschaft

Zitieren

APA

Shah, Z. H., Müller, M., Hübner, W., Wang, T. C., Telman, D., Huser, T., & Schenck, W. (2024). Evaluation of Swin Transformer and knowledge transfer for denoising of super-resolution structured illumination microscopy data. GigaScience. https://doi.org/10.1093/gigascience/giad109

Vancouver

Shah ZH, Müller M, Hübner W, Wang TC, Telman D, Huser T, et al. Evaluation of Swin Transformer and knowledge transfer for denoising of super-resolution structured illumination microscopy data. GigaScience. 2024. doi:10.1093/gigascience/giad109

BibTeX
@article{Shah2024,
  author = {Shah, Zafran Hussain and Müller, Marcel and Hübner, Wolfgang and Wang, Tung-Cheng and Telman, Daniel and Huser, Thomas and Schenck, Wolfram},
  title = {Evaluation of Swin Transformer and knowledge transfer for denoising of super-resolution structured illumination microscopy data},
  journal = {GigaScience},
  year = {2024},
  doi = {10.1093/gigascience/giad109}
}
Kostenlos starten →