If you search for volume rendering on this forum you will find lots of questions to give you ideas of where people have troubles. I strongly suggest you pick a specific problem to focus your work rather than starting with a general problem of volume rendering. Examples could be cone beam CT for dental planning or improving vessel visualization in the brain in MRA datasets. Perhaps someone here who has a particular use case can give you a problem to work on.