Intelligent Tutoring Systems for Adaptive Learning Pathways in Healthcare Training

Area of Science:

• Medical Education
• Artificial Intelligence
• Robotic Surgery

Background:

• Traditional surgical training is resource-intensive and inefficient.
• Robotic-assisted surgery (RAS) requires specialized, effective training methods.
• Personalized and scalable training solutions are needed for surgical education.

Purpose of the Study:

• To develop and evaluate an AI-powered Recommender System for personalized RAS training.
• To enhance surgical skill acquisition through adaptive task selection.
• To create a scalable and efficient training framework for robotic surgery.

Main Methods:

• Developed a two-step Recommender System: a data-driven decision base and a reinforcement learning (RL) decision algorithm.
• Utilized a synthetic dataset based on Item Response Theory Knowledge Tracing Model for simulation.
• Employed a graph-based knowledge tracing model to identify latent task structures.

Main Results:

• The graph-based knowledge tracing model effectively supported the decision base by revealing task relationships.
• Reinforcement learning within the decision algorithm enhanced the selection of optimal training tasks.
• The AI framework demonstrated potential for accelerating learning in simulated RAS training.

Conclusions:

• AI-powered Recommender Systems offer a promising approach for personalized and scalable RAS training.
• Integrating knowledge tracing and reinforcement learning improves adaptive task selection.
• Further research is needed for real-world implementation and optimization of the AI training framework.