Would it be nice to read electrical impulses without intrusion... The problem
as to why we can't do this currently no matter how scientifically useful is
that if you use external measurements of magnetic or electic fields there is
just too much noise. Remember, musculature also produces currents (hence the
ECG), not to mention the fact that your dealing with a great big, wet, volume
conductor. Then comes the problem of picking out individual signals. This is
hard enough if you're sitting on the nerve, much less five feet away.
However, as a fictional piece you might look into the work that's being done
with cell phones where the phones' transmission reflections off of the body are
sent back to the receiving station. I forget who exactly is doing this, but
their current direction was a sort of medic-alert function for cellular
customers. I think that they can get pretty good signals for cardiac
assessment. You could stretch this to get what you want if you could
auto-majically filter out everything but the "pain" signals. A bit of a
stretch, but it is fictive.