Physics-driven self-supervised learning for quantitative high-fidelity structured illumination microscopy
- 1Key Laboratory of Breast Cancer Precision Medicine, Institute of Biomedical Engineering, Kunming Medical University, Kunming 650500, China
Abstract
Structured illumination microscopy (SIM) enables rapid, long-term super-resolution (SR) imaging of live-cell dynamics. However, current SIM reconstruction methods, although achieving high structural fidelity, generally lack reliable intensity quantification. Here we present qHiFi-SIM, a physics-driven self-supervised learning framework for quantitative high-fidelity SIM imaging. By leveraging the wide-field (WF) image as a physical intensity reference and conventional SIM reconstruction as a structural continuity prior, qHiFi-SIM enables self-supervised training without ground-truth intensity data. It achieves structural fidelity comparable to HiFi-SIM while maintaining excellent intensity linearity at the SIM point-spread function scale (R2 > 0.986, mean intensity ratio near 1). The framework transfers across diverse SIM setups and typical samples, and directly improves the intensity linearity of other state-of-the-art algorithms without retraining. We demonstrate its utility for live-cell quantitative visualization of mitochondrial intensity dynamics and high-fidelity SIM-FRET functional imaging, offering a practical tool for SR structural visualization and quantitative functional imaging in live cells.
Related
- Quantitative 3D imaging of intact engineered heart tissues using inverted selective plane illumination microscopy
- Supervised 3D Virtual Fluorescence and Pseudo-H&E Staining of Fresh Bulk Tissues via Multimodal qOBM and Fluorescence Microscopy
- Interpretable computational SHG microscopy enables quantitative tissue phenotyping of cardiac remodeling in nonischemic heart failure
- Microscopy Image Dataset of pulmonary vessels for Quantitative assessment of fibrosis
- Learning from Error Predictions Rather Than Prediction Errors: A Perspective on Self-Supervised Learning in the Brain
BibTeX
@misc{tang2026physicsdriven,
title = {{Physics-driven self-supervised learning for quantitative high-fidelity structured illumination microscopy}},
author = {Y. Tang and Z. Luo and X. Zhu and W. Wang and X. Ge and M. Li and C. Chen and T. Chen and C. Chen and P. Xi and G. Wen},
howpublished = {bioRxiv},
year = {2026},
month = jun,
doi = {10.64898/2026.06.12.731778},
url = {https://www.biorxiv.org/content/10.64898/2026.06.12.731778v2},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - Physics-driven self-supervised learning for quantitative high-fidelity structured illumination microscopy
AU - Y. Tang
AU - Z. Luo
AU - X. Zhu
AU - W. Wang
AU - X. Ge
AU - M. Li
AU - C. Chen
AU - T. Chen
AU - C. Chen
AU - P. Xi
AU - G. Wen
PY - 2026
DA - 2026/06/12
PB - bioRxiv
DO - 10.64898/2026.06.12.731778
UR - https://www.biorxiv.org/content/10.64898/2026.06.12.731778v2
AB - Structured illumination microscopy (SIM) enables rapid, long-term super-resolution (SR) imaging of live-cell dynamics. However, current SIM reconstruction methods, although achieving high structural fidelity, generally lack reliable intensity quantification. Here we present qHiFi-SIM, a physics-driven self-supervised learning framework for quantitative high-fidelity SIM imaging. By leveraging the wide-field (WF) image as a physical intensity reference and conventional SIM reconstruction as a structural continuity prior, qHiFi-SIM enables self-supervised training without ground-truth intensity data. It achieves structural fidelity comparable to HiFi-SIM while maintaining excellent intensity linearity at the SIM point-spread function scale (R2 > 0.986, mean intensity ratio near 1). The framework transfers across diverse SIM setups and typical samples, and directly improves the intensity linearity of other state-of-the-art algorithms without retraining. We demonstrate its utility for live-cell quantitative visualization of mitochondrial intensity dynamics and high-fidelity SIM-FRET functional imaging, offering a practical tool for SR structural visualization and quantitative functional imaging in live cells.
ER -