I am looking at an optical coherence tomography scheme which is based on a michelson interferometer. The basic idea is that a laser beam is split by a beamsplitter, half of which goes into some tissue and the other half into the reference arm. The tissue backscatters some of the light, the light is recombined with the light from the reference arm and creates interference, which can be used to visualize the tissue to a certain depth.
Now I am trying to combine this scheme with second harmonic generation which usually (?) occurs when an electric field (such as a laser) enters a non-linear medium. What happens is that if the laser with frequency $\omega_0$ enters this medium, another beam will leave the medium which is a mixture of $\omega_0$ and $2\omega_0$ frequencies. These can be then seperated by some prism for example.
Now it is good (depends how you look at it) that some tissues like collagen or cartilage act as non-linear mediums. As such when one tries to look at them in an OCT scheme they will back-scatter not only light with the initial frequency $\omega_0$ but also light with frequency $2\omega_0$.
Scientists have proposed some clever set-ups to be able to harness and use this extra light with frequency $2\omega_0$ to gain more information.
Now my main question: What are some drawbacks of using this method (as in using this second light $2\omega_0$)? In all papers I've read it seems to be strictly superior to the classical OCT method.
The only limiting factor I can think of that it is only possible to use this second harmonic effect in certain tissues which act as non-linear mediums. And even here I am not sure whether this is truly a fundamental limit, hence the title. Could we change the properties of tissue to make them all non-linear if necessary?
So again: What other drawbacks does this method have such that it is not common practice? And maybe...what are some ways to solve/compensate for these problems?