.png){kind=icon}
.png){kind=icon}
Return
MediCARE: Medical Collaborative Agents REasoning over Interpretable Heterogeneous Graphs
.png){kind=icon}
OA
AI
A
A
V
M
G
A
G
D
V
DOI：10.1016/j.artmed.2026.103444.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We introduce the Predict →Interpret →Explain (PIE) paradigm, a novel approach to leveraging Large Language Models (LLMs) in the medical domain. • The work acknowledges and addresses the risk of LLM hallucinations, proposing a multi-agent collaboration protocol to mitigate this risk. • The paper utilizes graph data structures and Graph Neural Network (GNN) models for the medication recommendation task, demonstrating a robust method for processing complex, interconnected patient clinical data. • The work employs eXplainable AI (XAI) techniques to interpret the GNN model and uses LLMs to further process the predictions and their interpretations, providing comprehensive, accurate, and easy-to-read reports for medical specialists. • The proposed framework, MediCARE, is tested on the MIMIC-III database, demonstrating the impact of LLMs in providing accurate reports and improving prediction precision.
Keywords:
Large Language Models
Collaborative LLMs
Graph Neural Networks
Personalized medicine
eXplainable Artificial Intelligence
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
6.2
Papers:
2.5K
Citations:
7.8K
Organization
Cited Papers