“The SlideTutor Project - An Intelligent Medical Training System for Visual Diagnosis"

Abstract: Cognitive Tutors are adaptive, computer-based educational systems that provide individualized instruction by incorporating models of expert performance and dynamically building a unique student model for each user. They typically simulate real-world tasks - students work through problems or cases as the Tutoring System offers guidance, points out errors and organizes the curriculum to address the needs of that individual learner. Can cognitive tutors be effective in ill-structured domains such as medical diagnosis? Over the last two years we've developed a general framework for tutoring of visual classification problem-solving, and have used this framework to create SlideTutor - a model tracing Intelligent Tutoring System in Pathology. Many aspects of medical domains pose challenges for ITS design. One of these challenges is that there are no formal notations to be exploited in the interface. System designers must invent their own notations that make visible the intermediate reasoning steps in diagnostic classification. What kinds of representations are effective and how well are these representations accepted? During this talk I'll describe early results of our first formative evaluation of SlideTutor, in which we explored the effects of problem representation on performance gains, metacognitive gains and user experience.

Bio: Rebecca Crowley, MD, MS is the Principal Investigator of the SlideTutor project. She is an Assistant Professor (since 2001) at the Center for Pathology Informatics, University of Pittsburgh School of Medicine, and also has faculty appointments in the Intelligent Systems Program, University of Pittsburgh Faculty of Arts and Sciences and in the Medical Informatics Training Program, Center for Biomedical Informatics. Her training includes a residency in Anatomic Pathology, fellowship in Neuropathology, National Library of Medicine Fellowship in Medical Informatics, and a Master's degree in Information Science. Her research interests include: development and evaluation of intelligent medical training systems, medical knowledge representation and decision support, information extraction from medical free-text, and empirical studies of the development of visual diagnostic expertise. Her work has been funded by the National Library of Medicine and National Cancer Institute.