I agree with @kephale and @imagejan (after working on this for quite a bit now :D) that writing a whole new visualiser with VR support might be a little much for an undergrad project. Getting a basic version working is easy, but the devil is really in the details.
That being said, I at the moment have a very capable intern who has integrated new ambient occlusion and shadowing algorithms into scenery, and along these lines I could imagine feasible undergrad design projects for your students, such as:
- integration of an editor for multi-D transfer functions, which could be then adjusted with the Vive or Oculus controllers
- better interfacing with ImageJ, as @kephale suggested, e.g. to have all the ImageJ commands (e.g. for acting on volumetric data) available in a "floating" VR menu
- integration of non-photorealistic rendering, which can be very useful if you want to emphasize outlines, etc. in renderings
- (your idea here :D)
The only requirement for the student would be to be comfortable programming, and accept a steep learning curve. Everything graphics-related usually comes with a high amount of mathematics (mostly linear algebra, though).
I hope this was helpful, we're of course open for more discussion about topics, and so on