Enhancing Intelligent Triage with Large Language Models: A Comprehensive Evaluation and Optimization Study.

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare

Background:

• Intelligent triage systems are crucial for efficient patient management.
• Current triage methods face challenges in accuracy and resource optimization.

Purpose of the Study:

• To evaluate the effectiveness of large language models (LLMs) in intelligent patient triage.
• To enhance triage accuracy, efficiency, and healthcare resource allocation using LLMs.

Main Methods:

• Assessed various LLMs across diverse triage scenarios.
• Employed prompt engineering, parameter tuning, and workflow design for optimization.
• Developed customized LLM configurations for specific healthcare needs.

Main Results:

• Optimized LLM configurations significantly improved triage accuracy.
• LLMs demonstrated adaptability to different triage scenarios.
• Demonstrated potential for high-precision triage assistance in real-world healthcare settings.

Conclusions:

• Customized LLMs offer substantial improvements in patient triage accuracy.
• LLM integration can lead to enhanced patient care and optimized resource allocation.
• The study provides a framework for implementing LLMs in clinical triage workflows.