DiagnoCare: A Diagnostic Analysis System for Electronic Medical Records Combining Medical Knowledge Graph and Large Language Model

With the rapid development of healthcare technologies and the increasing demand for intelligent diagnostic systems, we present a novel approach that combines medical knowledge graphs and Large Language Models (LLMs) for enhanced Electronic Medical Record (EMR) analysis and condition warning. Our system integrates structured medical knowledge with advanced text analysis capabilities to accurately interpret and process complex EMR data. By leveraging LLMs and prompt learning, the system provides personalized examination suggestions, diagnostic analyses, and disease warnings. Experimental results on real-world datasets, including the MIMIC III database and anonymized hospital records from Fujian Province, China, demonstrate the system’s efficacy in delivering precise and actionable insights. Case studies further illustrate the system’s practical value and usability, highlighting its potential to significantly improve clinical decision-making and patient outcomes. This research offers substantial implications for the advancement of intelligent EMR systems and personalized healthcare.