Intelligent Tutoring System for Medical Students: Opportunities and Challenges

Medical education is constantly evolving in both approaches and content. Medical schools face the challenge of training students on various technological and practical skills. Traditional theoretical lectures need to be supported by adjusted learning approaches to ensure that future doctors master the diagnosis and classification of pathologies, such as diabetic retinopathy (DR).The present research work aims to develop an AI tutoring system-based prototype to support the diagnosis of diabetic retinopathy and improve methods of learning for medical students in ophthalmology. To capture the needs and expectations for such tools, we performed a series of 21 interviews with ophthalmologists. The analysis of the interviews leads to seven core requirements that support an AI-driven learning tool. We designed this approach as an active learning process. We also discuss an ongoing supported effort at utilizing common tools to make it easier for the learners to reach their expected skill levels and provide them with a rich learning experience. Our prototype provides new opportunities such as self-learning, simulation, and automatic assessments. Based on the analysis of doctors’ needs, our prototype is evaluated by comparing its results against the expected objectives expressed during the interviews.