# Is H&E Image-to-Spatial Transcriptomics Simpler Than It Looks?

arXiv: https://arxiv.org/abs/2609.32857 · HTML: https://mlsanity.com/paper/2609.32857/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 19 references that are in the map.
- First posted on arXiv 26 Sep 2026.

## Builds on (most cited first, 11 of 11)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [HEST-1k: A Dataset for Spatial Transcriptomics and Histology Image Analysis](https://mlsanity.com/paper/2406.16192/) (2024)
- [Spatially Resolved Gene Expression Prediction from H&E Histology Images via Bi-modal Contrastive Learning](https://mlsanity.com/paper/2306.01859/) (2023)
- [Scalable Generation of Spatial Transcriptomics from Histology Images via Whole-Slide Flow Matching](https://mlsanity.com/paper/2506.05361/) (2025)
- [SEPAL: Spatial Gene Expression Prediction from Local Graphs](https://mlsanity.com/paper/2309.01036/) (2023)
- [A Large-Scale Benchmark of Cross-Modal Learning for Histology and Gene Expression in Spatial Transcriptomics](https://mlsanity.com/paper/2508.01490/) (2025)
- [HistoPrism: Unlocking Functional Pathway Analysis from Pan-Cancer Histology via Gene Expression Prediction](https://mlsanity.com/paper/2601.21560/) (2026)
- [Teaching pathology foundation models to accurately predict gene expression with parameter efficient knowledge transfer](https://mlsanity.com/paper/2504.07061/) (2025)
- [HEXST: Hexagonal Shifted-Window Transformer for Spatial Transcriptomics Gene Expression Prediction](https://mlsanity.com/paper/2605.04682/) (2026)
- [HiST: A Hierarchical Sparse Transformer for Cross-Modal Spatial Transcriptomics Modeling](https://mlsanity.com/paper/2606.14251/) (2026)

## Abstract (arXiv)

Predicting spatial gene expression from routine H&E histology offers a scalable route toward spatial molecular profiling. Recent work has pursued increasingly sophisticated architectures to capture spatial context and richer expression structure. At the same time, simple estimators have shown strong performance in several studies, but what they already solve and where additional complexity is needed remain unclear. We study this behavior through the structure of prediction error under the mean-squared error (MSE) objective. Differences in average expression across genes can account for a substantial part of aggregate prediction performance, while a key unresolved error lies in recovering variation within each slide. Decomposing MSE into slide-level and within-slide components, we find that the within-slide component has lower residual-normalized parameter sensitivity in controlled neural experiments. This motivates Component-Guided Loss (CGL), which increases supervision of the within-slide component. CGL-Linear is a closed-form affine instantiation that achieves overall state-of-the-art performance across HEST-1k cohorts and gene-panel sizes. The same within-slide supervision improves existing neural models. These results suggest that substantial gains can come from aligning the training objective with prediction-error structure rather than increasing model complexity.