Scholarlink

Abstract <p>Distant metastasis (DM) is the primary driver of cancer-related mortality, and its clinical prediction remains challenging due to the lack of robust biomarkers. This study proposes a novel graph representation that effectively identifies discriminative morphological features from histopathological whole slide images (WSIs). By transforming high-resolution WSIs into topological graphs, the proposed method leverages graph neural networks (GNNs) to capture complex spatial dependencies and cellular organizations critical for metastatic progression. The study is evaluated on a large-scale pan-cancer dataset and demonstrates superior performance in distilling shared metastatic patterns across diverse malignancies. Furthermore, the cross-dataset robustness of this representation is validated by training on a specialized nasopharyngeal carcinoma cohort (TJ-NPC) and evaluating on independent public datasets. The results highlight the potential of computational pathology to provide scalable, objective risk stratification, offering a high accuracy tool for personalized clinical intervention.</p> Graphical Abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Pan-cancer Distant Metastasis Prediction Based on Graph Neural Network

  • 期刊论文

  • 发表日期：

    2026年5月11日

查看全文
  • Fengyun Zhang,
  • Qiangguo Jin,
  • Changming Sun,
  • Ruibing Chen,
  • Jie Geng,
  • Siqi Chen,
  • Wenrun Cai,
  • Xugang Sun,
  • Xiaofeng Liu,
  • Ran Su

摘要

Abstract

Distant metastasis (DM) is the primary driver of cancer-related mortality, and its clinical prediction remains challenging due to the lack of robust biomarkers. This study proposes a novel graph representation that effectively identifies discriminative morphological features from histopathological whole slide images (WSIs). By transforming high-resolution WSIs into topological graphs, the proposed method leverages graph neural networks (GNNs) to capture complex spatial dependencies and cellular organizations critical for metastatic progression. The study is evaluated on a large-scale pan-cancer dataset and demonstrates superior performance in distilling shared metastatic patterns across diverse malignancies. Furthermore, the cross-dataset robustness of this representation is validated by training on a specialized nasopharyngeal carcinoma cohort (TJ-NPC) and evaluating on independent public datasets. The results highlight the potential of computational pathology to provide scalable, objective risk stratification, offering a high accuracy tool for personalized clinical intervention.

Graphical Abstract