Digital case library: a resource for teaching, learning, and diagnosis support in radiology

Abstract

The authors compiled a digital case library, a database of cases for intracranial masses, that can be used as an electronic teaching file and image source for case-based teaching applications, electronic textbooks, and diagnosis support tools. The library is a relational database with a feature-coded image archive that is structured around relevant radiologic findings. Its index for coding image content is structured as a hierarchical image description index that uses the relational format. Rules that control the search direction within the library and that generate lists of diagnostic hypotheses for decision support tools are embedded within the database structure. Currently, the library consists of 200 cases and 1,100 images that present intracranial masses on radiographs, computed tomograms, magnetic resonance images, and angiograms. Each image in the library is indexed according to its radiologic content. The user may search for reference images that contain particular radiologic features by formulating image content-based queries. The hierarchical index of radiologic findings allows multilevel query formulation that depends on the user's level of experience.