Multi role ChatGPT framework for transforming medical data analysis

Area of Science:

• Artificial Intelligence in Medicine
• Medical Data Science
• Computational Biology

Background:

• The accuracy of large language models like ChatGPT in medical applications is under scrutiny.
• Existing methods for medical data analysis can be time-consuming and prone to errors.

Purpose of the Study:

• To introduce a novel Multi-Role ChatGPT Framework (MRCF) to improve ChatGPT's accuracy and efficiency in medical data analysis.
• To compare the performance of MRCF against traditional manual analysis and a singular ChatGPT model.

Main Methods:

• Developed the Multi-Role ChatGPT Framework (MRCF) by optimizing prompts, integrating real-world data, and implementing quality control.
• Evaluated MRCF's performance in interpreting medical data and identifying inaccuracies.
• Created two user-friendly databases for drug repositioning analysis using the MRCF.

Main Results:

• MRCF demonstrated significantly higher accuracy and fewer random errors compared to manual analysis.
• MRCF was over 600 times more time-efficient and one-tenth the cost of conventional manual annotation.
• The framework successfully facilitated efficient drug repositioning analysis through dedicated databases.

Conclusions:

• The Multi-Role ChatGPT Framework (MRCF) offers a substantial advancement in the accuracy and efficiency of AI-driven medical data analysis.
• MRCF provides a cost-effective and rapid solution for complex tasks like drug repositioning.
• This framework has broader implications for improving data analysis models across various professional fields.