The prospect is great, the process is painful, I do not know whether you are now self-study or professional direction is this, to be honest, image reconstruction, alignment, fusion are in large imaging equipment on the more important technology, if you can seriously learn something, if you go to GPS (GE, Philips, Siemens) to do an algorithmic engineer or something is quite a play, the domestic This technology is relatively backward, if you have the ability to cooperate with their own learning some software development, their own drumming up a post-processing demo program out, and then find some of the gold masters over to cast a little bit, may also be able to become, such as CuiTongZhe. If you go to graduate school in the U.S., it is relatively easy to publish papers, but the employment options are not as good as the popular machine learning and data mining directions. So, if you are interested in scientific research and like this direction very much, and you may be in academia in the future, you can do this. If you consider working on the industrial street, there are not many places to go, just a few big medical instrument manufacturing companies, I guess.