Research map / Graph learning / Knowledge and temporal graphs
Graph-Semantic Guided Learning for Virtual Immunohistochemistry Staining on Consecutive Histology Sections
Fanhao Qiu, Yangyang Zhang, Zhengxia Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add GSRM (graph semantic rectification) to the baseline (ablation) | GSGStain generator · MIST (ER) | DAB KL divergence: 3.935 -> 2.128 | Improved |
| add GSRM (ablation) | GSGStain generator · MIST (ER) | FID: 57.608 -> 45.372 | Improved |
| add dual-branch discriminator (DBD) (ablation) | GSGStain generator · MIST (ER) | FID: 57.608 -> 42.326 | Improved |
| add DBD alone (ablation) | GSGStain generator · MIST (ER) | DAB KL divergence: 3.935 -> 3.438 | No clear effect |
| add DBD on top of GSRM (ablation) | GSGStain generator · MIST (ER) | FID: 45.372 -> 38.292 | Improved |
| add GSRM on top of DBD (ablation) | GSGStain generator · MIST (ER) | DAB KL divergence: 3.438 -> 1.966 | Improved |
| use UNI2-h instead of ResNet50 for H&E node features (ablation) | GSRM graph · MIST (ER) | FID: 42.518 -> 38.292 | Improved |
| replace handcrafted IHC priors with UNI2-h deep features (ablation) | GSRM node features · MIST (ER) | FID: 38.292 -> 55.116 | Hurt |
| add IHC prior features to graph nodes (ablation) | GSRM node features · MIST (ER) | FID: 59.842 -> 38.292 | Improved |
| reduce generator from 6 to 4 ResNet blocks (ablation) | GSGStain generator · MIST (ER) | FID: 38.292 -> 42.172 | Hurt |
| increase generator from 6 to 9 ResNet blocks (ablation) | GSGStain generator · MIST (ER) | FID: 38.292 -> 39.481 | Hurt |
| use UNet-128 generator instead of ResNet-6blocks (ablation) | GSGStain generator · MIST (ER) | FID: 38.292 -> 40.049 | Hurt |
Closest papers by meaning
- OT-StainNet: Optimal Transport Driven Semantic Matching for Weakly Paired H&E-to-IHC Stain Transfer (AAAI 2025 · similarity 0.91)
- ODA-GAN: Orthogonal Decoupling Alignment GAN Assisted by Weakly-supervised Learning for Virtual Immunohistochemistry Staining (CVPR 2025 · similarity 0.90)
- Virtual Immunohistochemistry Staining for Histological Images Assisted by Weakly-supervised Learning (CVPR 2024 · similarity 0.89)
- SheafStain: Sheaf-Theoretic Schr\"odinger Bridge for Spatially and Biologically Coherent Virtual Staining (NeurIPS 2026 · similarity 0.88)
- BioX-CPath: Biologically-driven Explainable Diagnostics for Multistain IHC Computational Pathology (CVPR 2025 · similarity 0.86)
- Diffusion Generative Modeling for Spatially Resolved Gene Expression Inference from Histology Images (ICLR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
