OSCEMentor

Realistic, adaptive patient interview simulation

What it does

OSCEMentor enhances medical education by providing an immersive environment for practicing patient interviews and OSCEs (Objective Structured Clinical Examinations). We use the Gemini API to create dynamic, lifelike patient simulations. Administrators create detailed patient scenarios, and the AI generates realistic patient responses. The cases can target a wide range of conditions and skill levels.
During the interview, the Gemini API enables natural, context-aware conversations. As the user asks questions or provides information, the AI patient responds in real-time, adapting its answers and behavior based on the conversation flow. This creates a highly realistic and unpredictable interaction, much like a real patient encounter. At any point, the user can request feedback on their performance thus far via Gemini as well. As users progress, the system dynamically adjusts the complexity and nuances of the cases, ensuring a continuous challenge that matches their growing expertise.
OSCEMentor's core functionality of AI-powered patient simulations and interactions is operational, with the real-time feedback system functional but undergoing refinement. In progress is using Imagen to dynamically generate visual representations of patients as a user progresses, as well as a more comprehensive feedback/evaluation framework for administrators. This project will be used with real students soon.

Built with

  • Web/Chrome

Team

By

Dinesh Rai

From

United States