I think it's very hard to find the correct stoichiometry with these techniques, particularly with EDS that is not so sensitive to the low concentration. You can try to use some analytical technique, for example ICP.
I totally agree with Thomas and Marcella. EDX and XRF are not (in particular EDX) sensitive enough to build an exact stoichiometry from experimental data. I would probably go for ICP too. Digest your solid in hot, concentrated H2SO4 and then check with ICP the amount of Ti and Cu that you have. If you know that the specimen has high purity, no other dopants and the synthesis has been carried out in a very controlled atmosphere to avoid impurities, than the rest should be oxygen.