I believe the most critical aspect is that you want to avoid selecting against sequences rather than structures. This means, if you go by a standrard SELEX protocol, you will finish with a seuqence that is a reverse complement of your target RNA. That would be an antisense probe, but not an aptamer. So what you need is to introduce counterselection steps to remove such sequences, and/or change your target RNA sequnce while mainting it structure - I do not know if that is feasible for you.
I know that the Mayer lab in Bonn is expert in these things, so I suggest you contact Günter Mayer: