Scalable Generation of Spatial Transcriptomics from Histology Images via Whole-Slide Flow Matching

TL;DR

This paper introduces STFlow, a flow matching generative model that predicts spatial transcriptomics from histology images by modeling cell interactions and efficiently processing entire slides, significantly outperforming existing methods.

Contribution

STFlow is the first model to explicitly incorporate cell-cell interactions and utilize an efficient slide-level encoder for whole-slide spatial transcriptomics prediction.

Findings

Outperforms state-of-the-art baselines on HEST-1k and STImage-1K4M datasets.

Achieves over 18% relative improvement over pathology foundation models.

Effectively models joint gene expression distribution across entire slides.

Abstract

Spatial transcriptomics (ST) has emerged as a powerful technology for bridging histology imaging with gene expression profiling. However, its application has been limited by low throughput and the need for specialized experimental facilities. Prior works sought to predict ST from whole-slide histology images to accelerate this process, but they suffer from two major limitations. First, they do not explicitly model cell-cell interaction as they factorize the joint distribution of whole-slide ST data and predict the gene expression of each spot independently. Second, their encoders struggle with memory constraints due to the large number of spots (often exceeding 10,000) in typical ST datasets. Herein, we propose STFlow, a flow matching generative model that considers cell-cell interaction by modeling the joint distribution of gene expression of an entire slide. It also employs an efficient slide-level encoder with local spatial attention, enabling whole-slide processing without excessive memory overhead. On the recently curated HEST-1k and STImage-1K4M benchmarks, STFlow substantially outperforms state-of-the-art baselines and achieves over 18% relative improvements over the pathology foundation models.