Skip to content

Latest commit

 

History

History
55 lines (47 loc) · 2.23 KB

2024-01-15.md

File metadata and controls

55 lines (47 loc) · 2.23 KB
Raw

2024-01-15

Agenda

  • DICOM v4

    • over 150 people interested

    • Initial meeting today at 11AM CST - probably logisitcs planning

    • Brainstorm of DICOMv4 Wish List

      • DICOM is not friendly enough for people who need to use it
        • AI Developers
        • Image Viewer developers
        • Analytics
        • First exposure to DICOM is tough
      • Documentation is too complex / hard to understand
        • Must be easier for new people to understand
        • Need for good "entry level" documentation
      • More focused on current state of the art
        • Get rid of parts that are not or rarely used
        • Simplify (fewer compression codecs)
        • Use modern codecs (json, etc)
        • Browsers, cloud
      • More real world use cases (like IHE)
      • DICOM Standard Process
        • Nothing should be standardized without an open source implementation
          • FHIR has a maturity model 0-5. "gazelle"
        • DICOM "Police" - some of kind enforcement, validation, verification
      • Basics/building blocks should be 100% solid and slowly changed
        • Ability to display an image properly (and generate a displayable image)
        • 2D, 3D, 4D, Cine Clips
      • Green / eco friendly / env friendly

    • Brainstorm of what we want to KEEP from DICOM v3

      • What does it do well?
        • Information model is good (what is a CT image?)
        • Acknowledgement (Storage Commitment)

    • What does DICOM not do well

      • Workflow
      • duplication of data between DICOM and HL7 - these should work together better
      • DIMSE/networking
      • Platform - AuthZ / AuthC, message bus, logging / auditing
      • DICOM Study filtering (usually end up downloading entire study)
      • Does not help with defining what a procedure/study and what a sequence/series is
        • RSNA Radlexx trying to do this
        • Various companies working on this
        • Big problem for AI and Hanging Protocols

  • DICOM Structured Reporting TID 1500 Update

    • EU is trying to use TID1500 to serve up AI results (siemens, ge, phillips tried but failed?)
    • meeting last week - 40 people showed up
    • no easy solution with reporting - if we are going to do highly structured like SR, there is going to be thousands of templates
    • TID1500 does not handle "nesting" well - drop down selections