https://www.indiegogo.com/projects/beta-bioled-the-first-hand-held-blood-analyzerWell, I don't have guts ti say it's scam right away, but it is at least very suspicious.
As I have some background in light measurements and chemistry (spectroscopy, radiospectrometry) - claiming, that their setup consisting of few assorted LEDs spread over the spectrum without any further filtering plus light integrator provided better results than commercital absobrtion spectrometers is just bullshit in my opinion.
Even with basic low cost miniautre spectrometer modules you have FWHM in 5nm range at least - and they are trying to convince me, that they can do better with LEDs, which spectra is gaussian shaped, spreaded over tens of nanometers will provide better results ? I don't think so. Based on this measurement, I would be able to resolve between basic colors of sample and mabye little bit about its transmittance, but hell no somethning about chemical composition of sample.
Only semiconductor light source capable of producing narrow enough spectra is laser diode, but then you are stucked basically to a few points in whole spectra - how could you estimate composition based on that (yep, Raman scattering, I know, but it really doesn't seem they are that far - and you are still tied to full spectra measuring and cost and effort for propper light filtering added) ? You simply won't get away without using prism, or diffraction grating - or am I missing something important ?
Oh, and did I mentioned that most valuable data about chemical composition of sample in case of optical spectrometry does not lie in visible part of spectra, but UV and infra ?
I'd like to hear somebody else opinion about this.