USIGAN: Unbalanced Self-Information Feature Transport for Weakly Paired Image IHC Virtual Staining

2025-07-08

AI Summary

Under weak pairing conditions, cross-domain generation from H&E to virtual IHC staining suffers from pathological semantic inconsistency caused by spatial heterogeneity across adjacent tissue sections, impeding reliable one-to-many mapping. To address this, we propose a spatial-alignment-free generative framework. Its core innovations are: (1) modeling the joint distribution of H&E and IHC via unbalanced optimal transport (UOT) to suppress noise inherent in weakly paired data; (2) introducing a pathology-aware self-correspondence mechanism (PC-SCM) and UOT-based consistency constraint (UOT-CTM) to jointly construct semantically consistent cross-domain association matrices at both image-level and intra-group levels; and (3) integrating unbalanced self-information feature transfer to strengthen morphological–staining semantic coupling. Evaluated on two public datasets, our method reduces IoD by 23.6% and improves Pearson-R by 18.4%, significantly enhancing clinical interpretability and practical utility.

Technology Category

• Computer Vision: Medical and Biological Imaging
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Systems and Infrastructure for Web, Mobile and WoT: Web applications in cross-disciplinary domains and verticals such as mixed reality, smart cities, and digital health

Abstract

Immunohistochemical (IHC) virtual staining is a task that generates virtual IHC images from H&E images while maintaining pathological semantic consistency with adjacent slices. This task aims to achieve cross-domain mapping between morphological structures and staining patterns through generative models, providing an efficient and cost-effective solution for pathological analysis. However, under weakly paired conditions, spatial heterogeneity between adjacent slices presents significant challenges. This can lead to inaccurate one-to-many mappings and generate results that are inconsistent with the pathological semantics of adjacent slices. To address this issue, we propose a novel unbalanced self-information feature transport for IHC virtual staining, named USIGAN, which extracts global morphological semantics without relying on positional correspondence.By removing weakly paired terms in the joint marginal distribution, we effectively mitigate the impact of weak pairing on joint distributions, thereby significantly improving the content consistency and pathological semantic consistency of the generated results. Moreover, we design the Unbalanced Optimal Transport Consistency (UOT-CTM) mechanism and the Pathology Self-Correspondence (PC-SCM) mechanism to construct correlation matrices between H&E and generated IHC in image-level and real IHC and generated IHC image sets in intra-group level.. Experiments conducted on two publicly available datasets demonstrate that our method achieves superior performance across multiple clinically significant metrics, such as IoD and Pearson-R correlation, demonstrating better clinical relevance.

Problem

• Achieving cross-domain mapping between H&E and IHC images
• Mitigating spatial heterogeneity in weakly paired conditions
• Ensuring pathological semantic consistency in generated results

Innovation

• Unbalanced self-information feature transport for IHC staining
• UOT-CTM mechanism for correlation matrix construction
• PC-SCM mechanism for intra-group level consistency