There is this program that I have been working on which is supposed to interpret MRI images. More specifically, it will try to isolate out the different parts in an image as being a part of various structures. The first step is to get the program to identify what is body wall; once that is done, you can assume anything inside the body wall is part of some internal organ or structure. This is considerably more complicated a task than I initally anticipated. After a year and a half and two complete rewrite, I finally have an algorithm which can do the first part of the task reasonably well, which is to identify the body wall. I would put up a sample image, but blogger doesn't seem to want to let me upload. Oh well... back to work