Intelligent tutoring system for radiology as a part of the information exchange infrastructure

Abstract

We have designed a radiology instructional system that is based on intelligent tutoring paradigm and capitalizes on resources that are already available, such as the picture archiving and communication system (PACS), clinical information system, and electronic textbook. The Radiology Puzzler is designed to support the learning of radiological patterns through graphical case retrieval and intended for teaching visual recognition of intracranial lesions on radiological images. Using radiologic feature samplers, the user may define brain lesion characteristics and use visual query when searching for reference cases. A rule-based module generates a hierarchical list of diagnostic hypotheses and provides a feed-back to the user. Relevant cases that match the sampler's index are retrieved from the case library for comparison with the case in question.