aiqbench analysis
This research integrates knowledge graphs into large language models to enhance the interpretability and accuracy of disease predictions using electronic health records. By scaffolding chain-of-thought reasoning with specific medical paths, the approach yields clinically preferred rationales that outperform classical baselines while demonstrating strong zero-shot transfer capabilities. The study highlights a practical pathway for deploying lightweight instruction-tuned LLMs in high-stakes healthcare settings where explainability is critical.
Knowledge Graph Augmented Large Language Models for Disease Prediction
arXiv:2512.01210v4 Announce Type: replace Abstract: Electronic health records (EHRs) enable strong clinical prediction, but explanations are often coarse and hard to use for patient-level decisions. We propose a knowledge graph (KG)-guided chain-of-thought (CoT) framework for visit-level disease prediction on MIMIC-III. We map ICD-9 codes to PrimeKG, mine disease-relevant nodes and paths, and use these paths to scaffold temporally consistent CoT rationales, retaining only samples whose conclusions match observed outcomes. We fine-tune lightweight instruction-tuned LLMs (LLaMA-3.1-Instruct-8B and Gemma-7B) on two small cohorts (400 and 1,000 index visits) across ten PrimeKG-mapped diseases. Our models outperform strong classical baselines, reaching AUROC 0.66-0.70 and macro-AUPR 0.40-0.47. Without additional training, the models transfer zero-shot to the CRADLE cohort, improving accuracy from 0.40-0.51 to 0.72-0.77. In a blinded clinician study, KG-guided CoT rationales are consistently preferred for clarity, relevance, and correctness. Code is available at: https://github.com/JonathanWry/KG-guided-LLM-pipeline
Abstract:Electronic health records (EHRs) enable strong clinical prediction, but explanations are often coarse and hard to use for patient-level decisions. We propose a knowledge graph (KG)-guided chain-of-thought (CoT) framework for visit-level disease prediction on MIMIC-III. We map ICD-9 codes to PrimeKG, mine disease-relevant nodes and paths, and use these paths to scaffold temporally consistent CoT rationales, retaining only samples whose conclusions match observed outcomes. We fine-tune lightweight instruction-tuned LLMs (LLaMA-3.1-Instruct-8B and Gemma-7B) on two small cohorts (400 and 1,000 index visits) across ten PrimeKG-mapped diseases. Our models outperform strong classical baselines, reaching AUROC 0.66-0.70 and macro-AUPR 0.40-0.47. Without additional training, the models transfer zero-shot to the CRADLE cohort, improving accuracy from 0.40-0.51 to 0.72-0.77. In a blinded clinician study, KG-guided CoT rationales are consistently preferred for clarity, relevance, and correctness. Code is available at: this https URL
Artificial Intelligence (cs.AI)
arXiv:2512.01210 [cs.AI]
(or arXiv:2512.01210v4 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2512.01210 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 229 words
Related coverage
Building evidence-based knowledge bases from full-text literature for disease-specific biomedical reasoning
arXiv · Sep 9, 2026
CliQ-RRG: Clinical-Knowledge Guided Disease-aware Visual-Textual Alignment for QA-Style Radiology Report Generation
arXiv · Oct 12, 2026
Presentation: Platform Teams Enabling AI - MCP/Multi-Agentic Tools across Linkedin
InfoQ · Jun 5, 2026
AI in healthcare: applications and best practices
Databricks · Jul 24, 2026