Research map / Vision-language and multimodal / Human and embodied modeling
Transitive Representation Learning Enhances Histopathology Annotation
Moritz Schaefer, Zoe Piran, Nils Philipp Walter, Animesh Awasthi, Christoph Bock, Jure Leskovec, Zinaida Good
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 |
|---|---|---|---|
| train on I-G and G-T pairs with transitive alignment instead of bimodal I-T pretraining | SpatialWhisperer · PathoCell, Lizard | mean AUROC: PathoCell 0.566 -> 0.621; Lizard 0.642 -> 0.749 | Improved |
| train on I-G and G-T pairs instead of bimodal I-T QUILT-1M training | SpatialWhisperer · PanNuke cell type | mean AUROC: 0.713 -> 0.687 | Hurt |
| joint trimodal alignment instead of two-stage image->expression->cell-type pipeline | SpatialWhisperer · PathoCell, Lizard, PanNuke | mean AUROC: 0.550 -> 0.621; 0.662 -> 0.749; 0.599 -> 0.687 | Improved |
| joint alignment of image, expression and annotation instead of marker-gene-only bridge | SpatialWhisperer vs OmiCLIP · PathoCell | mean AUROC: 0.491 -> 0.621 | Improved |
| harmonize text label style of I-T data with G-T data (LLM rewrite) (ablation) | QUILT-1M labels · QUILT-1M image-text retrieval | AUROC: 0.645 -> 0.695 | Improved |
| use curated instead of uncurated QUILT-1M data in trimodal training (ablation) | SpatialWhisperer · PathoCell | mean AUROC: 0.609 -> 0.645 | Improved |
| add transitive data (trimodal) to subsampled task-matched data (ablation) | bimodal training · low-data regimes (G->T, I<->G retrieval, skin conditions) | AUROC | Improved |
| add transitive data at full task-matched data (ablation) | bimodal training · full data | AUROC | No clear effect |
| add I-T supervision to T-G + G-I bridge training (ablation) | SpatialWhisperer · PanNuke 19-class tissue-of-origin | AUROC: 0.586 -> 0.649 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PLIP — PLIP: Language-Image Pre-training for Person Representation Learning (NeurIPS 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- STimage-1K4M: A histopathology image-gene expression dataset for spatial transcriptomics (NeurIPS 2024 · similarity 0.89)
- HEST-1k: A Dataset For Spatial Transcriptomics and Histology Image Analysis (NeurIPS 2024 · similarity 0.88)
- M2OST: Many-to-one Regression for Predicting Spatial Transcriptomics from Digital Pathology Images (AAAI 2025 · similarity 0.88)
- Diffusion Generative Modeling for Spatially Resolved Gene Expression Inference from Histology Images (ICLR 2025 · similarity 0.88)
- Semantic and Visual Crop-Guided Diffusion Models for Heterogeneous Tissue Synthesis in Histopathology (NeurIPS 2025 · similarity 0.88)
- Towards Scalable Context-Aware Single-Cell Spatial Transcriptomics Prediction from Histology Images (NeurIPS 2026 · similarity 0.88)
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.
