Beyond Traditional Diagnostics: Transforming Patient-Side Information into Predictive Insights with Knowledge Graphs and Prototypes

2025-12-09

AI Summary

To address the long-tailed distribution bias and insufficient interpretability in disease prediction from patient-side data (e.g., demographics and self-reported symptoms), this paper proposes an interpretable prediction framework integrating medical knowledge graphs with disease prototypes. Methodologically: (i) it constructs clinically interpretable disease prototype representations; (ii) it introduces knowledge graph–guided contrastive learning to enhance discriminative capability for rare diseases; and (iii) it leverages large language models to generate diagnosis explanations that are medically sound, patient-specific, and semantically aligned with patient narratives. Experiments on real-world clinical datasets demonstrate that our method significantly improves prediction accuracy for tail-end diseases (+5.2% F1-score), while explanation quality achieves high clinical consensus (89% inter-rater agreement among domain experts). The framework thus advances both predictive performance and model transparency, offering a reliable foundation for early clinical intervention.

Technology Category

• Machine Learning: Transparent, Interpretable, Explainable ML
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Humans and AI: Explainable AI (XAI) for Human Understanding

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Predicting diseases solely from patient-side information, such as demographics and self-reported symptoms, has attracted significant research attention due to its potential to enhance patient awareness, facilitate early healthcare engagement, and improve healthcare system efficiency. However, existing approaches encounter critical challenges, including imbalanced disease distributions and a lack of interpretability, resulting in biased or unreliable predictions. To address these issues, we propose the Knowledge graph-enhanced, Prototype-aware, and Interpretable (KPI) framework. KPI systematically integrates structured and trusted medical knowledge into a unified disease knowledge graph, constructs clinically meaningful disease prototypes, and employs contrastive learning to enhance predictive accuracy, which is particularly important for long-tailed diseases. Additionally, KPI utilizes large language models (LLMs) to generate patient-specific, medically relevant explanations, thereby improving interpretability and reliability. Extensive experiments on real-world datasets demonstrate that KPI outperforms state-of-the-art methods in predictive accuracy and provides clinically valid explanations that closely align with patient narratives, highlighting its practical value for patient-centered healthcare delivery.

Problem

• Predict diseases from patient-side information like demographics and symptoms
• Address imbalanced disease distributions and lack of interpretability in predictions
• Enhance predictive accuracy and provide medically relevant explanations for reliability

Innovation

• Integrates medical knowledge into disease graph
• Uses contrastive learning for long-tailed diseases
• Generates patient-specific explanations with LLMs