Resolution of skin auto-fluorescence for non-invasive application

Auto-fluorescence spectrum indicates the cell and tissue state during physiological, pathological process, distribution and chemical-physical properties. However, the spectrum is difficult to distinguish that result largely from strong overlap. This article present the genetic algorithm to decompose the overlapping spectrum. To do so, the simulated spectra can be performed to confirm our feasible algorithm. The skin auto-fluorescence spectra can then be obtained from the normal human skin with 337 nm excitation light source. Via the proposed methods, the nicotinamide adenine dinucleotid (NADH) and flavin adenine dinucleotide (FAD) accounted for about 96.6% and 3.3% respectively of the overall spectrum. The developed algorithm can be widely applied to achieve qualitative and quantitative analysis for overlapping spectra.