OSCEai dermatology: augmenting dermatologic medical education with Large Language Model GPT-4

Area of Science:

• Medical Education Technology
• Artificial Intelligence in Healthcare
• Dermatology Training

Background:

• Traditional medical education faces challenges in providing diverse clinical case exposure.
• Integrating technology can enhance practical skills training for healthcare professionals.
• Visual-based specialties like dermatology require specialized training tools.

Purpose of the Study:

• To demonstrate the integration of large language models (LLMs) into medical education.
• To enhance dermatology trainees' history taking and management skills using an OSCE-like format.
• To provide a novel, interactive learning tool for medical education.

Main Methods:

• Utilized GPT-4, a large language model, to generate diverse, realistic skin cancer role-play scenarios.
• Incorporated pre-existing, evidence-based images across various skin tones.
• Developed an OSCE-like format for simulated patient encounters.

Main Results:

• The LLM-driven tool successfully generated realistic scenarios for dermatology training.
• Personalized feedback was provided to learners based on their performance.
• The tool demonstrated potential for improving diagnostic skills and patient-centered care.

Conclusions:

• LLM-driven educational tools can effectively supplement traditional medical teaching methods.
• This approach offers a scalable solution for exposing trainees to a wider range of clinical cases.
• The innovation supports the development of patient-centered and culturally sensitive care.