Hierarchical Multi-Scale Graph Learning with Knowledge-Guided Attention for Whole-Slide Image Survival Analysis

2026-02-26

AI Summary

This work proposes a hierarchical multi-scale knowledge-aware graph network to address the limitations of existing whole-slide image survival analysis methods, which often neglect spatial architecture or rely on static handcrafted graphs, thereby failing to effectively model multi-scale, hierarchical tissue relationships. The proposed approach constructs dynamic local cell-level graphs to aggregate neighboring image patches and integrates regional features at the global level, jointly leveraging coarse-grained contextual and fine-grained structural information for prognosis prediction. By incorporating spatial locality constraints into the dynamic graph structure, the method enables knowledge-guided multi-scale feature fusion, overcoming the spatial modeling limitations inherent in conventional multiple instance learning frameworks. Evaluated on four TCGA cohorts, the model achieves an average 10.85% improvement in concordance index, with statistically significant patient risk stratification (log-rank p < 0.05).

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

We propose a Hierarchical Multi-scale Knowledge-aware Graph Network (HMKGN) that models multi-scale interactions and spatially hierarchical relationships within whole-slide images (WSIs) for cancer prognostication. Unlike conventional attention-based MIL, which ignores spatial organization, or graph-based MIL, which relies on static handcrafted graphs, HMKGN enforces a hierarchical structure with spatial locality constraints, wherein local cellular-level dynamic graphs aggregate spatially proximate patches within each region of interest (ROI) and a global slide-level dynamic graph integrates ROI-level features into WSI-level representations. Moreover, multi-scale integration at the ROI level combines coarse contextual features from broader views with fine-grained structural representations from local patch-graph aggregation. We evaluate HMKGN on four TCGA cohorts (KIRC, LGG, PAAD, and STAD; N=513, 487, 138, and 370) for survival prediction. It consistently outperforms existing MIL-based models, yielding improved concordance indices (10.85% better) and statistically significant stratification of patient survival risk (log-rank p<0.05).

Problem

• whole-slide image
• survival analysis
• multi-scale
• spatial hierarchy
• cancer prognostication

Innovation

• hierarchical graph learning
• multi-scale integration
• knowledge-guided attention
• dynamic graph
• whole-slide image analysis