WSI-GT: Pseudo-Label Guided Graph Transformer for Whole-Slide Histology

ICLR 2026 Conference Submission17788 Authors

Keywords: graph transformers, pseudo-labeling, digital pathology, whole-slide images, patch classification, over-smoothing mitigation

TL;DR: WSI-GT is a pseudo-label guided graph transformer for classifying patches in 100k×100k whole-slide histology images, combining graph convolutions and intra-class attention to achieve state-of-the-art accuracy

Abstract:

Whole-slide histology images (WSIs) can exceed 100k × 100k pixels, making direct pixel-level segmentation infeasible and requiring patch-level classification as a practical alternative. However, most approaches either treat patches independently, ignoring spatial and biological context, or rely on deep graph models that oversmooth, leading to loss of critical tissue details.

We present WSI-GT (Pseudo-Label Guided Graph Transformer), a simple yet effective architecture that addresses these challenges. WSI-GT combines a lightweight local graph convolution block for neighborhood feature aggregation with a pseudo-label guided attention mechanism that preserves intra-class variability and mitigates oversmoothing. To cope with sparse annotations, we introduce an area-weighted sampling strategy that balances class representation while maintaining tissue topology.

WSI-GT achieves a Macro F1 of 0.95 on PATH-DT-MSU WSS2v2, improving by up to 3 percentage points over tile-based CNNs and by about 2 points over strong graph baselines. It further generalizes well to the Placenta benchmark and standard graph node classification datasets, highlighting both clinical relevance and broader applicability. These results position WSI-GT as a practical and scalable solution for graph-based learning on extremely large images.

Primary Area: learning on graphs and other geometries & topologies

Submission Number: 17788