This is a question for David Clunie @dclunie, but this came up in the past, and instead of sending email I am posting this here, for the benefit of the community. I will try to remember to post such questions of general interest in the public forum going forward.
In one of the models we are using, we have a need to encode segmentation of the structure that has laterality but also has additional positional modifier - “lower lobe of left lung” (and of course other combinations of those modifiers).
In the past, for coding “lung”, we would specify laterality as a modifier code. But when I search for this structure in SNOMED, I seem to see only the codes that already include the modifiers (see below).
Should we use those pre-coordinated codes and part with our prior practice, or there is another way? Can this be clarified in the DICOM CIDs somehow? Currently, CID 6110 Lung Anatomy Finding or Feature only has the code for “Lobe of lung”.
DICOM Segmentation value multiplicity for the Segmented Property Type Modifier Code Sequence Attribute is 1, so we cannot have multiple modifiers to say “Left” and “Lower” at the same time (see Segmented Property Type Modifier Code Sequence Attribute – DICOM Standard Browser and C.8.20.2 Segmentation Image Module).
Truly unfortunate, ChatGPT is below mediocre in helping pick SNOMED codes (at least when I tried to get help): ChatGPT cannot help with SNOMED (yet?) - Google Docs. We need a customization “text2snomed”, to follow on the earlier “text2cohort” work!