GCUNet: A GNN-Based Contextual Learning Network for Tertiary Lymphoid Structure Semantic Segmentation in Whole Slide Image

TL;DR

This paper introduces GCUNet, a GNN-based model that effectively captures long-range contextual information for improved TLS semantic segmentation in large whole slide images, outperforming state-of-the-art methods.

Contribution

The paper presents GCUNet, a novel GNN-based network that aggregates long-range context for TLS segmentation, along with four new datasets for benchmarking.

Findings

GCUNet achieves at least 7.41% higher mF1 than SOTA methods.

Four new TLS segmentation datasets are introduced, totaling 826 WSIs.

GCUNet effectively integrates context and detail for accurate segmentation.

Abstract

We focus on tertiary lymphoid structure (TLS) semantic segmentation in whole slide image (WSI). Unlike TLS binary segmentation, TLS semantic segmentation identifies boundaries and maturity, which requires integrating contextual information to discover discriminative features. Due to the extensive scale of WSI (e.g., 100,000 \times 100,000 pixels), the segmentation of TLS is usually carried out through a patch-based strategy. However, this prevents the model from accessing information outside of the patches, limiting the performance. To address this issue, we propose GCUNet, a GNN-based contextual learning network for TLS semantic segmentation. Given an image patch (target) to be segmented, GCUNet first progressively aggregates long-range and fine-grained context outside the target. Then, a Detail and Context Fusion block (DCFusion) is designed to integrate the context and detail of the target to predict the segmentation mask. We build four TLS semantic segmentation datasets, called TCGA-COAD, TCGA-LUSC, TCGA-BLCA and INHOUSE-PAAD, and make the former three datasets (comprising 826 WSIs and 15,276 TLSs) publicly available to promote the TLS semantic segmentation. Experiments on these datasets demonstrate the superiority of GCUNet, achieving at least 7.41% improvement in mF1 compared with SOTA.