Benchmarking Pathology Foundation Models for Spatial Domain Understanding
2026-05-25 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial Intelligence
AI summaryⓘ
The authors created SpaPath-Bench, a new way to test how well pathology AI models understand the layout of different tissue types in medical images. Instead of just checking if the models can predict diseases, they looked at whether the models can recognize meaningful regions and their spatial relationships using both images and gene data. They tested many models and methods on a large dataset and found that different training styles help the AI learn different parts of tissue structure. Their work helps improve future AI tools for analyzing tissue images more precisely.
Pathology foundation modelsWhole slide imagesSpatial transcriptomicsSpatial domain identificationRepresentation learningTissue architectureUnsupervised learningComputational pathologyPretraining paradigmsBenchmarking
Authors
Bokai Zhao, Yiyang Zhang, Yuanchi Zhu, Hanqing Chao, Long Bai, Tai Ma, Minfeng Xu, Ming Song, Tianzi Jiang
Abstract
Pathology foundation models (PFMs) have emerged as a core approach for learning transferable representations from whole slide images (WSIs), and they are typically benchmarked through downstream clinical endpoints. While such task level evaluations are indispensable, they offer limited insight into what the representations themselves encode, particularly whether PFM embeddings can distinguish meaningful tissue regions and capture their spatial relationships. We present SpaPath-Bench, a representation level benchmark designed to diagnose spatial representation capability in PFMs. SpaPath-Bench formulates spatial domain identification (SDI) on paired whole slide image and spatial transcriptomics (ST) data as a diagnostic task. It curates 42 public paired WSI and ST slides, enables large scale evaluation across 19 encoders and seven SDI methods, and measures partition quality using three complementary criteria: unsupervised spatial coherence, transcriptomics referenced agreement, and expert referenced agreement. Across 83K runs, SpaPath-Bench reveals that different pretraining paradigms capture distinct aspects of tissue spatial architecture, and it provides practical guidance for building the next generation of spatially aware computational pathology models. Code and data pipelines are publicly available at https://bokai-zhao.github.io/SpaPath-benchboard/.