PaLM Conf Minutes 2020-July-08

Attendees

Name	Email
Gunter Haroske	haroske@icloud.com
Mary Kennedy	mkenned@cap.org
Raj Dash	r.dash@duke.edu
Rikki Merrick	rikimerrick@gmail.com
Alessandro Sulis	Allesandro.sulis@crs4.it
Ralf Herzog	ralf.herzog@roche.com
David de Mena Garcia	david.mena.exts@juntadeandalucia.es
Megumi Kondo	megumi.kondo.sakura.japan@gmail.com
Dan Rutz	drutz@epic.com
Kevin Schap	kschap@cap.org
David Beckman	dbeckman@epic.com
Francesca Frexia	Francesca.frexia@crs4.it
Jim Harrison	james.harrison@virginia.edu
JD Nolen	jdlnolen@gmail.com
Nick Haarselhorst	Nick.Haarselhorst@philips.com
Nick Jones	thatwsiguy@gmail.com
Ian Gabriel	Ian.gabriel@leicabiosystems.com
Filip Migom	Filip.migom@mips.be
John Hargett	jhargett@epic.com
Mandel Mickley	Mandel.mickley@leica-biosystems.com

Next call is August 12, 2020

• Digital Pathology update:
  • Document has been sent to Mary Jungers
  • Sharing the spreadsheet of open issues
    • Still need to have to update the wiki for this profile
    • Also put the comparison spreadsheet here: https://wiki.ihe.net/index.php/APW_Image_Acquisition
    • Harmonization proposal for HL7
      • Need to submit for code for identifier type of “Image Display Sort Order” (IDSO); it would be applicable for the assigning authority that creates it, so there could be more than one for an image – Riki to do
      • Slides have to be properly labeled correctly to convey the cut order (7th cut off the block); that is important, but the sorting is done on string, so that can be hard to do in the system natively
        • Sorting based on how the slides are created is most important
          • This should be created in DICOM Supplement 122
        • Also, there may be rules in protocols that govern sort orders
        • It is good to be able to have a way to create a sort order for the image (from the acquisition manager) first
        • There is also sorting by the users at a later point in time
        • DICOM has presentation states – this may be needed for the long-term image storage to cover application and end-user sorting – up to vendors to implement
        • What do we do if there are changes of the sort order?
          • This is in LAB-80, but not in the image display back report, but this should probably be also part of the DICOM metadata; we are just providing a mechanism to share
          • We may want to have image view trails / annotations and sort order changes may be stored there so that the case comes back up the same way
          • Limited use here: display identifier is just giving the sort order for the new images to be acquired – all else is out of scope
          • What if the sort order was changed – how would the acquisition manager know that the change has happened?
          • That should be covered in the evidence creation transactions in new profile
          • Comment from Nick Jones: To add a comment: I think it's important to differentiate the physical abstraction change vs. a display sort order. Nick H is right that labs vary a lot on how they describe these; one lab might say B2-4 for a slide, (the 4th cut on a second block on a second part of a case) while another lab describes that as 2-B-4, or 2-B-D, etc. That should be recognized as different descriptions of assets, not necessarily sort order for viewing (which is what the viewer wants.)
            • When to use ORL^O34
              • Error in the handling of the acquisition modality
                • Cannot use the order message, system not ready
                • use LAB-82 – as soon as you start the work
• SET Update
  • Updated document
    • Added diagrams
    • Fixed typos
    • As pathology is moving into digital world
      • For current slides we have manual process / physical to ensure pathologists know that the case is NOT complete
      • Need to be sure that in the digital world we need to be sure the case is not signed out
      • Imaging vendors and LIS vendors need to have a way to figure out, when all the physical assets are complete vs when all the digital assets are available
      • SET transaction telling the acquisition manager that it has been received at the stainer for example
      • Vendors can implement SET for this purpose
      • This is just the framework for trigger event driven data exchange; that the vendors can then build application on these profiles, for example the application can be build using SET to give information on status of digital assets
  • Specimen derivation
    • All other events in SET are driven by other profiles, but this one is not yet fully defined
      • Will the current message structure support the digital asset development?
      • Message structure includes the order information for the processing step, resulting in the derived specimen
        • We need to update the OO change request to this message structure
      • Nick Jones comment:
        • I may have some ideas on additions for events to add to these tables after I talk with the DICOM group (i.e. modeling the reduction of a specimen by its creation of a child specimen). But I'll follow up with you later. The basics here look great to me though; it solves some problems I hadn't figured out yet
        • It might just be that events like "Specimen Procedure Step successfully produced a derived specimen" get associated with both the parent and child specimens in various systems
  • Need to update the event reason table
  • Almost ready for public comment
• Next F2F:
  • Plan to postpone planning