Digital Staining with Knowledge Distillation: A Unified Framework for Unpaired and Paired-But-Misaligned Data

2025-04-14

AI Summary

High-quality paired training data are scarce in digital histological staining—particularly unpaired data or spatially misaligned paired data. Method: This paper proposes a two-stage unsupervised digital staining framework based on knowledge distillation. It introduces a novel light-enhancement–then-colorization generative architecture, a learnable spatial alignment module explicitly modeling pixel-wise correspondences between adjacent tissue sections, and a hybrid no-reference loss incorporating NIQE guidance to eliminate reliance on ground-truth images. Contribution/Results: To our knowledge, this is the first method unifying treatment of both unpaired and misaligned-paired scenarios without requiring pixel-level alignment annotations. Evaluated on a self-collected dataset and the WBC dataset, it achieves PSNR gains of 2.1–3.8 dB and a 19.6% NIQE reduction over SOTA methods, with significantly improved fidelity in cellular morphology and localization—demonstrating strong clinical translatability.

Technology Category

• Computer Vision: Medical and Biological Imaging
• Machine Learning: Calibration & Uncertainty Quantification
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Staining is essential in cell imaging and medical diagnostics but poses significant challenges, including high cost, time consumption, labor intensity, and irreversible tissue alterations. Recent advances in deep learning have enabled digital staining through supervised model training. However, collecting large-scale, perfectly aligned pairs of stained and unstained images remains difficult. In this work, we propose a novel unsupervised deep learning framework for digital cell staining that reduces the need for extensive paired data using knowledge distillation. We explore two training schemes: (1) unpaired and (2) paired-but-misaligned settings. For the unpaired case, we introduce a two-stage pipeline, comprising light enhancement followed by colorization, as a teacher model. Subsequently, we obtain a student staining generator through knowledge distillation with hybrid non-reference losses. To leverage the pixel-wise information between adjacent sections, we further extend to the paired-but-misaligned setting, adding the Learning to Align module to utilize pixel-level information. Experiment results on our dataset demonstrate that our proposed unsupervised deep staining method can generate stained images with more accurate positions and shapes of the cell targets in both settings. Compared with competing methods, our method achieves improved results both qualitatively and quantitatively (e.g., NIQE and PSNR).We applied our digital staining method to the White Blood Cell (WBC) dataset, investigating its potential for medical applications.

Problem

• Reducing need for aligned stained-unstained image pairs
• Enabling digital staining via unsupervised deep learning
• Improving cell target accuracy in generated stained images

Innovation

• Unsupervised deep learning for digital staining
• Knowledge distillation with hybrid non-reference losses
• Learning to Align module for misaligned data