Creation of Radiology Teaching Content with STELLA-A STandardized Electronic Learning Library and Application...

Area of Science:

• Radiology Education Technology
• Medical Informatics
• Open-Source Software Development

Background:

• Traditional radiology teaching files face challenges with efficiency, obsolescence, and lack of standardized metadata.
• Existing systems are often vendor-specific, limiting interoperability and long-term usability.

Purpose of the Study:

• To develop and implement STELLA (a STandardized Electronic Learning Library and Applications), a vendor-neutral, open-source web application for radiology teaching case management.
• To overcome limitations of existing teaching file systems by providing robust tools for image collection, organization, annotation, and sharing.

Main Methods:

• STELLA was deployed as a containerized application on clinical servers, integrating with PACS via a 'send' button to retrieve DICOM images and metadata.
• Users are required to apply structured metadata tags (e.g., RadLex terms for subspecialty, anatomy, diagnosis) and can utilize customizable templates for extended data.
• Features include pseudonymization of protected health information, default case sharing, and support for custom projects, worklists, and multimedia capture.

Main Results:

• In 31 months, 131 users created 3467 teaching cases from 2791 patients across 17 subspecialties.
• Musculoskeletal (1793 cases) and Chest (535 cases) were the most represented subspecialties.
• Over 592 diagnoses and 67 core anatomy terms were recorded, with organized projects and worklists developed for systematic learning.

Conclusions:

• STELLA has been successfully developed and deployed, demonstrating rapid growth in case collection and utilization for teaching and mastery learning.
• The application's ontology-based data labeling and rapid case saving facilitate interoperability, enabling potential inter-institutional sharing for education, research, and AI development.