New framework GraphDx enhances medical diagnosis with cost-aware LLM knowledge graphs

Researchers have developed GraphDx, a novel framework designed to improve sequential diagnosis in medical settings. This system utilizes Large Language Models (LLMs) to construct Medical Diagnosis Knowledge Graphs (MDKGs) that are sensitive to both diagnostic relevance and cost. GraphDx employs three specialized agents—Perception, Reasoning, and Decision—to systematically gather information, score evidence, and plan diagnostic steps while minimizing expenses. Experiments on MedQA and MIMIC-IV datasets demonstrated significant improvements in diagnostic success rates, achieving 79-93% accuracy while simultaneously reducing testing costs by 20-54%. AI

IMPACT This framework could lead to more efficient and cost-effective medical diagnostic processes by leveraging LLMs for structured knowledge representation and reasoning.

RANK_REASON The cluster describes a new research paper detailing a novel framework for medical diagnosis. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Decision Agent
• DeepSeek-V3
• GraphDx
• Kimi K2
• Llama 3.3
• Medical Diagnosis Knowledge Graphs
• MedQA
• MIMIC-IV
• Perception Agent
• Reasoning agents in a dynamic world: The frame problem

• paper
• product
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →