Physics-Informed Deep Learning Model for Cross-Modality Super-Resolution in Fluorescence Microscopy

2026-07-23

AI Summary

This work addresses the limitation of purely data-driven approaches in cross-modal fluorescence microscopy super-resolution, which often violate the physical laws governing optical imaging. To overcome this, the authors propose a physics-informed generative adversarial network that integrates the microscope-specific point spread function (PSF) as a prior for translating confocal microscopy images to stimulated emission depletion (STED) super-resolution images. For the first time, the PSF is explicitly embedded into the training objective of the generative model, significantly enhancing both structural fidelity and physical plausibility of the synthesized images. Experimental results demonstrate that the proposed PSF-guided model outperforms non-PSF baselines in terms of structural accuracy, local bias control, and frequency-domain consistency, yielding outputs that more closely resemble real STED images.

Abstract

Cross-modality image translation offers a route to super-resolution fluorescence microscopy from low-resolution images while reducing phototoxicity and instrumentation demands. However, purely data-driven models can produce visually plausible outputs that are inconsistent with optical image formation. Here, we propose a physics-informed generative adversarial network for confocal-to-STED image translation that incorporates microscope-specific point spread function information into the training objective. Simulated and experimentally measured PSFs were evaluated using a limited paired confocal-STED dataset of TOM20-labeled mitochondria in human primary M2 macrophages acquired across different experimental days. Performance was assessed using reference-based and non-reference-based image-quality metrics, together with complementary frequency- and distribution-sensitive analyses. The no-reference metrics probed physics-relevant image properties, including spatial-frequency content, contrast, and signal-to-noise behavior. PSF-guided models improved structural fidelity, reduced local deviations, and achieved closer agreement with STED references than non-PSF baselines, particularly in frequency-domain analyses. These results demonstrate that optical priors can improve the structural fidelity and physical plausibility of generative microscopy models for cross-modality super-resolution imaging.

Problem

• cross-modality super-resolution
• fluorescence microscopy
• physics inconsistency
• image translation
• optical image formation

Innovation

• physics-informed deep learning
• cross-modality super-resolution
• point spread function (PSF)
• generative adversarial network
• fluorescence microscopy