LabVIEW Vision is one of the more difficult and "advanced" topics to tackle. I had a lot of programming experience, including 4-5 years of maintaining and re-writing a Real-Time LabVIEW routine to do behavioral testing of human subjects (a sound-localization study), when a colleague asked me to help him save 5-10 second videos for an animal behavior study -- it took the two of us several months to work out how IMAQdx worked, and learning to deal with the 4-dimensional aspect of video images.
You need to get someone with LabVIEW Vision experience to look at your LabVIEW code and "get you started", including finding where the elusive ROIs are being stored and used. If you are in a University environment where such expertise might be found, reach out. Without seeing your code, and having a chance to talk with you to fully understand what you are trying to do, progress will be difficult (and slow).
Bob Schor