Page History

...

I think what SDTM has not addressed with imaging results is that when you look at the images produced by a procedure, and if your task it to look for the occurrence of a suspected object, what should be considered as the anatomical location of the TEST since you are looking at a 2-dimentaional image/picture. My take is that the images are still representative of, and are about a specific section or part of the body, it allows you to view the entirety of a section or part of the body. In other words, you are looking for a suspected object within a section/part of the body that is made visible to you by the diagnostic procedure. So applying this logic, I outlined the modeling for coronary occlusion data.

...

Lastly, i suppose you could represent aneurysms, arterial dissection, coronary occlusion all as clinical events and use the FACE structure, although i think they are findings about cardiovascular system so CV is a better place for them. However using FACE I have arrived to the same conclusion, see below. What's interesting about the concept map below is that it further shows how anatomical location values vary in --LOC depending on the SDTM class the data go into. As you summarized:

1. Anatomical focus of an intervention - at which part of the body an intervention is being made. (PR)

2. Anatomical manifestation of an event - the part of the body which shows a sign of the event occurring. (CE)

3. Anatomical object of an observation - about which part of the body is the observation being made. (FA/CV)

This shows me that a single --LOC variable for three classes, is an issue.

...