Mmm we're looking to incorporate some kind of UV laser into our cancer screening device. The UV range <400nm provides the most useful information, but since we're trying to keep the cost low so that it's affordable for underdeveloped countries, it doesn't look like UV will be doable
The laser diodes are coupled to 100-200um optical fibers which are pressed against the skin of a patient. another fiber adjacent to that probe fiber is then connected to a photodiode to measure the reflectance/scattering.
We are also looking into LED's but LED's being a noncoherent light source, it's hard to get a usable amount of light into the fiber and back to the photodiode.
If there are no UV laser diodes available you'd end up with a tripled yag laser, not the cheapest. You could look into nitrogen lasers, but those have a horrible beam quality and probably won't get much light in the fiber.