Lung-R1: A Knowledge Graph-Guided LLM for Pulmonary Diagnostic Reasoning

2026-06-10

AI Summary

This study addresses the “knowledge-to-diagnosis gap” between general medical knowledge and personalized reasoning from electronic medical records (EMRs) in pulmonary disease diagnosis. To bridge this gap, the authors construct LungKG, the first comprehensive lung diagnosis knowledge graph encompassing 15 entity types and 112 relation types, and introduce a knowledge graph–guided training paradigm for large language models. By integrating knowledge-constrained chain-of-thought reasoning with reinforcement learning, they develop Lung-R1-14B, a specialized large model for pulmonary diagnosis. The model achieves a score of 4.3583 on the EMR Diagnosis task, significantly outperforming the strongest baseline by +0.1476, and establishes new state-of-the-art performance across three benchmarks: Choice, Pulmonary-QA, and EMR Diagnosis.

Abstract

Diagnosing pulmonary diseases requires integrating heterogeneous evidence amid phenotypic variability and cross-disease overlap. Although large language models (LLMs) have shown progress on pulmonary knowledge question answering (QA) and information-processing tasks, reliable pulmonary diagnosis requires patient-specific, relation-aware reasoning over electronic medical record (EMR) evidence rather than isolated knowledge recall. We define this gap between pulmonary knowledge and case-level diagnostic reasoning as the Pulmonary Knowledge-to-Diagnosis Gap. To address it, we introduce LungKG, the first structured pulmonary knowledge graph for diagnostic knowledge organization and record-grounded reasoning. LungKG contains 59,038 nodes and 164,308 edges across 15 entity types and 112 relation types, serving as both a reusable pulmonary knowledge resource and the foundation for LungKG-guided model adaptation. Built on LungKG, we propose Lung-R1, a LungKG-guided pulmonary LLM trained through KG-constrained reasoning-chain construction and KG-guided reinforcement learning. In a 20-system evaluation, Lung-R1-14B achieves state-of-the-art performance across Choice, Pulmonary-QA, and EMR Diagnosis, reaching an EMR Diagnosis score of 4.3583 and surpassing the strongest non-Lung-R1 baseline by 0.1476 points. These results demonstrate the value of LungKG-guided training for EMR-based pulmonary diagnosis.

Problem

• pulmonary diagnosis
• knowledge-to-diagnosis gap
• electronic medical record
• diagnostic reasoning
• phenotypic variability

Innovation

• Knowledge Graph
• Large Language Model
• Diagnostic Reasoning
• Pulmonary Disease
• Reinforcement Learning