Self-Explaining Hypergraph Neural Networks for Diagnosis Prediction

• DOI: 10.48550/arxiv.2502.10689
• Publication Date: 2025-02-15
• Authors: Leisheng Yu; Yanning Cai; Minxing Zhang; Xia Hu
• Keywords: Hypergraph

The burgeoning volume of electronic health records (EHRs) has enabled deep learning models to excel in predictive healthcare. However, for high-stakes applications such as diagnosis prediction, model interpretability remains paramount. Existing prediction with intrinsic often assign attention weights every past or hospital visit, providing explanations lacking flexibility and succinctness. In this paper, we introduce SHy, a self-explaining hypergraph neural network model, designed offer personalized, concise faithful that allow interventions from clinical experts. By modeling each patient unique employing message-passing mechanism, SHy captures higher-order disease interactions extracts distinct temporal phenotypes personalized explanations. It also addresses the incompleteness EHR data by accounting essential false negatives original record. A qualitative case study extensive quantitative evaluations on two real-world datasets demonstrate superior performance over existing state-of-the-art models.