Sexual neuroscience is one of the most under-researched areas in the human sciences. A quick search of PubMed (the international database of medical research) shows that we know more about the neuroscience of hiccups than we do about orgasm.
Part of the problem is practical. fMRI scanners, some of the most useful and popular tools in cognitive neuroscience, involve lying in a tube while scanning takes place and need the head to be completely still. Add the fact that you're being watched by neuroscientists and none of this makes for relaxed coupling, or even self-stimulation.
Nevertheless, pioneering researchers have tried to tackle the problem with other methods. A Dutch team led by Gert Holstege used an alternative form of brain scanning, known as Positron emission tomography or PET, to study orgasm in men. PET doesn't have the same cramped conditions as fMRI allowing for a bit more sexual freedom in the lab.
They still had some considerable problems to overcome though, not least of which was timing an orgasm to occur during a predefined time-slot, during which brain activity could be monitored. PET requires a radioactive tracer to be injected into the bloodstream, and although the radiation is very weak, it's best to use only as much as necessary. This means the intuitive approach of continuous scanning and waiting for the pop of the cork is just not feasible.