Conceptual framework and documentation standards of cystoscopic media content for artificial intelligence

Area of Science:

• Urology
• Medical Informatics
• Data Science

Background:

• Clinical documentation of cystoscopy involves visual and textual data.
• Secondary use of visual cystoscopic data for education and research is limited by inefficient data management.

Purpose of the Study:

• To design a conceptual framework for standardized cystoscopy documentation.
• To ensure data management, annotation, and utilization align with FAIR principles.

Main Methods:

• Developed a framework with data, annotation, and utilization management sections.
• Proposed a Swiss-cheese model for quality control.
• Defined infrastructure for FAIR data principles implementation.
• Applied data sharing scenarios for research and education.

Main Results:

• Successfully implemented the framework adhering to FAIR principles.
• Created a cystoscopy atlas for an educational web portal.
• Made 68 qualitative videos and annotation data available for AI projects (frame classification, segmentation).

Conclusions:

• The proposed framework standardizes visual documentation storage.
• Enables FAIR data for educational purposes and artificial intelligence research in urology.