Stream: implementers
Topic: DICOM tags in FHIR
Santosh Bhavani (Aug 31 2020 at 18:07):
I'm working on a measure aimed at reducing radiation dosage in CT exams. In order to be compatible with CQL engines, I need to represent several DICOM tags within a FHIR resource. I was thinking of adding custom codes (e.g. kVp, CTDIvol) to Observation. Is this a reasonable approach?
Lloyd McKenzie (Aug 31 2020 at 19:13):
@Elliot Silver
Elliot Silver (Aug 31 2020 at 22:13):
I suggest looking at the work done on Cancer reporting (https://build.fhir.org/ig/HL7/fhir-breast-radiology-ig/index.html) by @Richard Esmond , @Brian Bialecki and others. I think that is the most in-depth interaction between FHIR and DICOM that has been documented, and could likely serve as inspiration. @Christopher Lindop and @Jonathan Whitby are current co-chairs of the Imaging workgroup and could probably speak to other initiatives.
In the past, the work group tried to clearly delineate between FHIR and HL7 content, and reduce duplication (and the potential for divergent values). I'd be hesitant to encode "too" much from the DICOM Dose SR into FHIR. Instead, allow a FHIR resource to point to the SR as a source of "more" information.
All that said, yes, custom observation codes seem to make sense. I suggest not using the DICOM element tag value, since your codes won't include the full path to the element. Instead, make up codes and provide mappings from your code to DICOM elements. I suspect if this is going to go beyond a fairly closed environment, @Jonathan Whitby and @Christopher Lindop would like to get the WG involved.
Elliot Silver (Aug 31 2020 at 23:09):
Also, I believe @Abderrazek Boufahja is looking at a similar issue.
Santosh Bhavani (Sep 01 2020 at 03:06):
@Elliot Silver Thank you for the thoughtful response! I'll look into the past work for inspiration. I'd love to reference DICOM resources but I'm constrained by existing tooling that operates only on FHIR resources. This will be submitted to CMS so maybe it would make sense to engage more folks in the WG.
Abderrazek Boufahja (Sep 02 2020 at 09:09):
Hello, I am working with the DICOM WG20 for the description of the dose summary through FHIR, however I worked also in the presentation of Radiation Dose SR structures in FHIR custom rsources. I think we are going through the same path.
Brian Bialecki (Sep 15 2020 at 15:13):
@Santosh Bhavani I have been working on these representations in FHIR. I have a few POC items I have been working on. These are in github here. https://github.com/ACRCode/dsi-standards-examples/tree/master/CDE%20%2B%20AIM This is something I have been looking to expand past my work with Stanford to the vendor community. If you wish to talk please let me know.
Abderrazek Boufahja (Dec 01 2020 at 17:05):
@Santosh Bhavani Here is the prepared draft on the Dose Summary on FHIR : https://confluence.hl7.org/display/IMIN/Dose+Summary+on+FHIR
Any comment is welcome.
Last updated: Apr 12 2022 at 19:14 UTC
