FPM2Stain Net: Physics-Guided Single-Shot Super-Resolution and Multi-Modal Virtual Staining for Digital Histopathology

We propose FPM2Stain Net, an end-to-end computational pipeline that integrates physics-guided super-resolution with deep learning–based virtual staining to enable high-resolution, multi-modal digital histopathology. The first stage, Single-Shot Fourier Ptychographic Microscopy (SS-FPM), reconstructs high-resolution amplitude and phase maps from a single 4× brightfield image using a self-supervised ResNet-U-Net guided by a differentiable point spread function (PSF) model. The second stage employs a multi-task conditional generative adversarial network (cGAN) enhanced with wavelet-based spatial–frequency fusion and perceptual supervision to synthesize clinically relevant staining modalities, including H&E, DAPI, LAP2, and panCK. Extensive experiments on both simulated and real tissue datasets demonstrate that FPM2Stain Net outperforms conventional FPM, GAN-based, and diffusion-based methods in both reconstruction fidelity and staining accuracy. Remarkably, the synthesized virtual stains not only preserve fine structural details but also enable downstream analyses such as cell segmentation and biomarker quantification with higher accuracy than real 40× microscope images. These results confirm that FPM2Stain Net achieves a greater-than-10× effective resolution enhancement over the original 4× input and provides a fast, scalable, and cost-effective alternative to chemical staining, with broad applications in digital pathology, multiplex imaging, and point-of-care diagnostics.