Application of Artificial Intelligence Generated Content in Medical Examinations

Area of Science:

• Medical Education Technology
• Artificial Intelligence in Healthcare

Background:

• Large language models and artificial intelligence generated content (AIGC) present new possibilities for medical examination question development.
• Current methods for effectively utilizing AIGC in designing medical questions are not well-established.
• Paper-based examinations have inherent limitations that AIGC could potentially overcome.

Purpose of the Study:

• To explore the effective utilization of AIGC for designing medical examination questions.
• To provide a streamlined instruction for generating various question types using AIGC.
• To highlight the potential of AIGC in transforming medical examinations and education.

Main Methods:

• Development of a streamlined instruction for generating medical questions using AIGC.
• Focus on creating multiple-choice questions, case study questions, and video questions.
• Inclusion of manual review to ensure accuracy and quality of AIGC-generated content.

Main Results:

• AIGC demonstrates rapid response capabilities, high efficiency, and effectiveness in mimicking clinical realities for question generation.
• A structured approach for AIGC-based question design has been provided.
• Manual review is confirmed as a necessary step for quality assurance.

Conclusions:

• AIGC offers a promising avenue for enhancing the creation of medical examination questions, improving efficiency and realism.
• Future integration of technologies like retrieval augmented generation and multi-agent systems will further advance AIGC capabilities in this domain.
• The evolution of AIGC is expected to significantly improve medical examination preparation and contribute to cultivating skilled medical professionals.