PaLM Conf Minutes 2019-Mar-20

From IHE Wiki
Jump to: navigation, search

Recording

Attendees

Name Name
Megumi Kondo Alessandro Sulis
Raj Dash Francesca Frexia
Francois Macary JD Nolen
Kevin Schap Francesca Vanzo
Tom Oswell Nick Hasselhorst
Dan Rutz Ralf Herzog
Mikael Wintell Ian Gabriel
Uwe Horchner Jim Harrison
David Beckman Gabor Garanyi

Minutes

The next IHE PaLM call will be April 10, 2019.

SET

  • Alessandro sent updated Vol 2 version last week, includes the mapping of the SET data elements to the HL7 fields and specifically the meta data for each event – please review before 4/8
  • SET related CR will be presented to OO on April 18th call and hopefully then we have some feedback
  • Need to move the public comment due date to June 2019 – Riki to update Mary Jungers

F2F meeting in Tokyo

  • Need list of hotels
  • RSVP list – Mary to send out – need to know npo later by mid April; remote attendance is supported
  • Will start agenda topic collection, too

TMA Status

  • Dan or Filip no on

Digital Pathology

  • Update from Sunday meeting
    • 10 folks in room, 3 folks on phone, had scanner vendors and FDA rep and new pathologist from France and Venezuela
    • Google Drive link to profile metadata spreadsheet: https://drive.google.com/open?id=12i6hoEjX-h0k11vbg2VE0-jTSI_m42NcfFTlTQMhSH8
    • Open issues identified on Sunday:
      • Model, where actors retain the images for extended period of time, before it is committed to image manager / archive (for QC), or the image manager is used multiple times – important not to have images not accessible to folks that need it
      • Multiple imaging steps for same glass slide ID – scan same slide at 20x and 40x – the scanner won’t know which one is which, since the glass slide ID that is the only ID accessible to the scanner from the physical asset.
        • May be need to add in multiple communications to inform all scanners of the active order(s)
        • Scanners would have to attach the IWOS to the glass slideID; need to know the condition of the slide (stain then de-stain and re-stain – but there would be temporal aspect, so the first order would already be done – so orders cannot remain in the system, once they are completed
      • What to do, when more than 3 channels are needed – DICOM limitation (example would be fluorescent stains 4 to 5, but sometimes up to a dozen signals)
        • Could we create new DICOM image for each channels (currently each filter is applied and then you take a snapshot – that is the representation of what is seen for the same region of interest) – can do, but loses a lot of dynamic imaging functions that is otherwise available in DICOM = this will be a work around; will bring this up at next DICOM F2F
      • What to do with parent specimen IDs for QC and PT – how do you attach them to the case?
        • On-slide control – need to indicate that control worked as part of the report
        • Separate QC specimen
        • PT models a patient specimen, so would be handled the same way
      • Riki to still develop the multiple SPM approach to show the examples vs using free text in SPM-14
      • Please review the google spreadsheet and make comments in it
      • Response includes the OBR-3 = Study instance ID – there should be only one for all the images OO just added a Filler grouper number (ORC-38), while it is globally unique, it is not sure all scanners generate the DICOM study UID – so should this be the image UID? No that is currently in OBX-5
        • What is the studyID – lung and lymphnode = entire case, related to accession number, so only 1 per case
        • DICOM IDs:

DICOM E-R Model.png

Suggested DICOME Structure.png

  • So we have Case – Study – series – image – at each level we have a UID
    • In DICOM there is only 1 image /series, so the study ID groups multiple images together
    • SOP instance ID is one level of the image
  • Do you need to have filler order number?
  • When you store the level in the stack, you create the entre image from that stack – each level has a different orderID in the PACs
  • What is the smallest object in DICOM?
    • The levels – the IDs are assigned to each object by the scanner
    • For the manager, need to only need to know that the WSI scanning process is completed, so one of these – the base level image ID may be all that is needed for that

Next meeting will be F2F in Warwick on 4/10 – with call in information IHE PaLM topics will be in the PM