Well... an opto coupler requires power, so you need to add still more complexity, and yet another power source for isolation. Then of course, how do you test that the opto has not failed, since the operating condition and the usual broken condition look the same? And given the cost of an opto and the cost of a PIC these days are not that far apart, you are sort of into diminishing returns. So IMHO, an opto is not a good solution in this instance.

The previous poster's twisted pair recommendation is very good. A CLC filter would be good too, but again, a bit of overkill for this application, I think. RC filter is plenty, unless you are worried about radiating electromagnetic interference from your circuit with the antennae you are stringing all over the place... in that case a CLC filter is more bidirectional, and will help prevent you causing problems for your neighbors. Your board layout and clock choices and enclosure will all play a role here too - EMC in non-trivial.