Compound terms getting a grammar hit?

Aside from just turning off this grammar rule, which I think we probably shouldn’t as it does find some real things, what options do we have for preventing hits on compound terms like “Stent thrombosis”?

In this paragraph, there are 3 hits, but we also get hits for “stent diameter”, etc.
Stent thrombosis was defined as occlusion of the stented venous segment occurring at any time following stent placement. Stent migration was defined as position change of a properly sized venous stent, with displacement of the stent outside of the intended treatment segment after the index procedure. Stent migration was determined with regard to a reference anatomic structure.

It is possible to use a pair of non word characters e.g. [ and ] or || and create a general ‘immunize rule’ in the disambiguator to remove any hit on word overlapping this area.
But… since part of the sentence is ‘mutilated’ with those characters, other rules might be influenced…
And the rule would have to be re-added after every release.

hmm. Can we define phrases (eg “stent thrombosis”) in some way that they supersede other grammar rules? Mangling the text doesn’t sound great…

I can’t think of a way to do that in LT, but you could ignore these matches when interpreting the API result.

Yeah, definitely an option.