If you want it to work out of the box then it will only work with the process of the mib RFC2790. (see image attached).
If that doesnt suit ur needs u would have to do some develops. I have done any of this yet, but i have a few ideas on how to achieve that could be a guideline for you to research.
option 1) Use a watch to monitor the process, whenever the process goes down generate an event. Using an event condition filter the time when it was trigger. If the event was generated in a maintenance period do nothing, otherwise generate an alarm.
The problem about using watchs is that the mib tables usually display the process that ARE running, so if a process stop running it will go out of the mib table. Consequently, u wont be able to generate an alarm. I think that if u use this option u will need to make one watch per process, which can be very annoying. There is probably an easier workaround to this issue but u would have to search for it.
option 2) The rfc2790 process works fine because whenever u add a new monitor, it generates a new model, so u can put that new model into maintenance. If it was up to me i would search for the file that triggers the add button on the "rfc2790 process monitoring" and i will try to copy it, but referencing the other mib tables instead of the rfc2790.