Efficient single-cell gene prediction from tissue images with context
Towards Scalable Context-Aware Single-Cell Spatial Transcriptomics Prediction from Histology Images
Computer Vision and Pattern Recognition
Summary
Measuring gene activity in tissue cells is usually slow and expensive. The authors developed a method called CELLO that predicts gene activity from microscope images of tissue stained with common dyes, without needing costly gene measurement. Their method looks at each cell individually while keeping its nearby surroundings in mind to improve accuracy. They show that CELLO is faster and more precise than previous approaches on diverse tissue samples.
What this means in practice
- •For hospital data teams: Produce detailed gene activity maps from routine tissue histology slides to support diagnostics without additional molecular testing.
- •For digital pathology software developers: Integrate efficient single-pass feature extraction for all cells in histology images to enable cell-level molecular predictions with local spatial context.
Authors
Zijun Gao, Chunbin Gu, Jinxi Xiang, Xiangde Luo, Pheng-Ann Heng
Abstract
Predicting gene expression from H&E-stained histology images offers a scalable alternative to costly spatial transcriptomics, yet most existing methods operate at the spot level, where signals from multiple cells are aggregated and critical cellular heterogeneity is obscured. Extending this paradigm to single-cell resolution is non-trivial. Naively applying pathology foundation models faces a scale mismatch: their patch-level representations mix multiple cells, whereas per-cell cropping or resizing distorts morphology and removes local context. Conversely, segmentation-based models without strong pretrained visual encoders often lack the morphological representation capacity needed for accurate molecular prediction and inherit errors from imperfect cell boundary masks. Here, we present CELLO, an efficient end-to-end framework that performs a single pathology foundation model forward pass per image and uses grid sampling to extract location-specific features for all cells simultaneously. We further introduce a distance-decay cross-attention module that refines each cell representation using spatially biased local morphological context. Using 52 public Xenium-H&E pairs from HEST-1k that span 12 organs and approximately 10 million cells, CELLO improves the average predictive accuracy over the evaluated baselines while reducing the mean whole-slide inference time compared to DeepSpot2Cell, a 14.0x speed-up on average that excludes upstream cell segmentation. Our work establishes a scalable foundation for single-cell gene expression prediction from H&E images.