We designed a tool that recognizes the “fingerprint” of each pharmaceutical compound when light is transmitted through an authentic sample at wavelengths in the ultraviolet (UV) light spectrum. This is the wavelength where the active pharmaceutical compound is reactive to light and is called the Baseline Spectral Fingerprint (BSF). We then leveraged the BSF through comparative assessment to infer quality of any random field sample and developed an algorithm to compensate for the randomness of field preparation using local sourced reagents.
We tested 125 samples of random substances prepared by a third party clinical pharmacologist. 20 were authentic samples while the remainder were prepared so that each sample was visually indistinguishable from the authentic samples but were either counterfeit or low quality (right medication, wrong concentration). The results yielded 1 false negative and 0 false positives.