Projects:RegistrationDocumentation

From NAMIC Wiki
Jump to: navigation, search
Home < Projects:RegistrationDocumentation

Back to ARRA main page

Use Case Library

Development Efforts

Reference Manual

  • The reference manual will contain detailed descriptions of each parameter and each control element within the slicer registration module. The description should help the user understand what exactly that function/parameter does and if/how useful it will be for their specific registration problem.
  • Each entry will have one short description (that could also serve as tooltip) and one longer, more technical explanation.
  • Preferred formats: Slicer Wiki, maybe PDF, cross-linking required
  • Draft Registration Reference Manual

User Manuals & Tutorials

  • The user manual will contain a systematic overview of registration functionality within 3DSlicer. formats: Slicer Wiki, PowerPoint.
  • Turtorials will be case oriented and always demonstrate a particular task or feature
  • Video Tutorials: showcase of specific workflow
  • Background Tutorials, explaining the basics of registration, formats: PowerPoint. Minimal understanding of the inner workings of a registration optimization algorithm is essential to understand and judge the results obtained and obtainable.
  • Topics for background tutorials:
    • the main components: transform, similarity function, optimization, interpolation
    • coordinate systems: physical vs. image space, RAS vs. LPI etc.
    • relevant image meta-data: image/axis orientation, image/CS origin, voxel size, dynamic range
    • overview of scenarios and their different challenges: image pairings, DOF, multi-modal, intra/inter-subject etc.
    • how to evaluate a match: tools & concepts
    • common mistakes to avoid: inappropriate DOF, overly flat similarity metric, CS inconsistencies, FOV discrepancies, wrong interpolation, insufficient search (sample points, multi-scale, DOF scale-space)
    • Troubleshooting guide: insufficient match - what next? Parameter modification, DOF change, initial alignment assist, fiducial help, ROI masking (e.g. skull stripping)
  • Tutorial Resources



Links

Bundled Registration and Tests