I think just distal to the retinaculum is best as there is a nice potential space in the synovial sheath to access.
If you inject right under the retinaculum it is very tight.
We scanning the short axis so we can see the liquid flow into both tendon sheaths (APL / EPB). If we see it is only going into one sheath you are dealing with a patient with seperate sheaths and we advance the needle into the second sheath and fill that as well.
It is worth identifying the superficial branch of the radial nerve so you don’t damage it. It is always close to the injection site.
I have had some radiologists inject the whole volume into the tendon sheath which I think is great and others will reserve a small volume and inject it outside the sheath over the retinaculum which I also think is fine.
As for CTS there is certainly patients where the clinical presentation is classic and the ultrasound is negative. I see this in acute settings and also chronic settings and also during pregnancy. I am never surprised by this as the ultrasound can not assess the nerve function so the nerve can look anatomically normal , however it is either under or over active with its function leading to the symptoms .
Remember :
“Treat the patient, not the imaging”
Steve