Cross-Modal Knowledge Distillation from Spatial Transcriptomics to Histology

TL;DR

This paper introduces a cross-modal distillation method that transfers spatial transcriptomics information to histology images, enabling tissue niche identification without transcriptomic data at inference.

Contribution

It presents a novel framework that leverages paired spatial transcriptomics and histology data to improve tissue niche detection in histology-only models.

Findings

Distilled models outperform morphology-based baselines in niche structure agreement.

The approach recovers biologically meaningful cell-type neighborhoods.

Framework works across multiple tissue types and disease contexts.

Abstract

Spatial transcriptomics provides a molecularly rich description of tissue organization, enabling unsupervised discovery of tissue niches -- spatially coherent regions of distinct cell-type composition and function that are relevant to both biological research and clinical interpretation. However, spatial transcriptomics remains costly and scarce, while H&E histology is abundant but carries a less granular signal. We propose to leverage paired spatial transcriptomics and H&E data to transfer transcriptomics-derived niche structure to a histology-only model via cross-modal distillation. Across multiple tissue types and disease contexts, the distilled model achieves substantially higher agreement with transcriptomics-derived niche structure than unsupervised morphology-based baselines trained on identical image features, and recovers biologically meaningful neighborhood composition as confirmed by cell-type analysis. The resulting framework leverages paired spatial transcriptomic and H&E data during training, and can then be applied to held-out tissue regions using histology alone, without any transcriptomic input at inference time.