PaLM Conf Minutes 2022-September-14
Jump to navigation
Jump to search
Attendees
Riki Merrick | Jason Lowder |
Raj Dash | Jim McNulty |
Dan Rutz | Rob Rae |
Alessandro Sulis | Mary Kennedy |
Jim Harrison | Kevin Schap |
Ruben Fernandes | Mandel Mickley |
Will Humphrey | Megumi Kondo |
Ralf Herzog | David Beckman |
Gianluca Pavan |
Next IHE PaLM call: Oct 12, 2022
Discussion
- Agenda review – no new items added
- Minutes – All posted minutes have been approved by general consent.
- Telemedicine white pater has been published and is in public comment until September 25. Any comments will be reviewed at the October PaLM call.
- HL7 / DICOM general questions
- See IHE PaLM listserv posting from Ruben 9.13.22
- How should block, slide, digital asset be distinguished?
- Currently we are discussing using specimen (SPM) and identifying the parent and using role to indicate which specimen information is “historical” -i.e., not submitted
- Each scan has a different ID
- We are currently using a pattern for the identifier to identify the block – cut – slide etc.
- Epic is not planning on extending the accessionID but will add in the DICOM ID into the model for each scan
- Phillips also has an additional image ID
- Can we identify the glass slice (using SAC) and would we do the same for the digital slide?
- Stains would be using SPM-6 (Additives) – had discussed that H&E stain (in both DICOM and HL7) that this would be 2 substances. We need to add these as codes to HL7 (might be harder in SNOMED, as these are substances vs H&E staining is a procedure). In FHIR we have specimen.processsing where we can use the procedure code from SCT = 104210008 H&E Stain (Procedure).
- There may be different protocols for H&E staining between labs; over time this can affect staining color. This is not currently covered or captured by the lab. This may be important for the future when using AI, etc.
- In the specimen DAM we have used a reference to the protocol ID so that the steps used can be documented elsewhere.
- This is similar to the harmonization field being discussed in other lab domains.
- Additive field is not designed to capture all the steps of a procedure, just the substances used.
- Using snapshot to indicate stain – unstained (empty SPM-8) and re-stained
- DICOM file has specific attributes that we need to get across in HL7
- Next Step:
- Identify all the use cases we need to document for AI
- Raj said the CAP AI committee can work on these use cases. Then, we can see what attributes are needed.
- Digital Pathology
- Waiting for someone to lead the Evidence Creation profile
- Have reached out to DICOM WG26 Co-Chairs, but no answers since they are currently prepping for their connectathon at Path Visions.
- How to deal with DICOM web
- DPIA may need to be updated for that
- No work on the DICOM 26 annotations supplement currently.
- Raj working on CR for DPIA
- For DPIA we could also check Query By Parameter (QBP); could help some of the scanner and LIS vendors.
- Plan for SET and DPIA Testing
- Need to create the message profiles
- Alessandro working on creating profile in IGAMT for SET – needs to also build it for DPIA (may be easier, since they are using existing message structures) – Alessandro will work on both – goal is to review during October 12 PaLM call.
- Need test cases written up for each of these profiles:
- Which to use and how many to include in a test plan for the 2023 Connectathons (US in February? and Europe in April)
- Next step for the LAW CR: send out to the group for voting
- Concordance Profile – no one from this item on call
- X-eHealth – no one from the item on call
- Next call is October 12, 2022