Adaptation of principal component analysis for continuous laser-induced fluorescence spectra decomposition. Krf excimer laser excitation examples

摘要

Laser-induced fluorescence (LIF) spectra of biological tissues spectra have continuous nature and can be statistically analyzed using principal components (PC) method. The disadvantage of this method is that PCs are abstract functions, alternating, giving both positive and negative contribution to the spectrum that it is impossible for the real fluorophores. In present paper the PC analysis was adapted to spectroscopy tasks by composing positive linear combinations of PCs corresponding to the peaks of the smallest width with the largest ratio of the spectral maximum to its integral across the measured wavelength range. This method was applied for the analysis of previously obtained LIF spectra of the intact teeth and teeth with caries, and spectra of aortal graft before and after decellularization. The representation of the principal components in the form of peaks with the smallest possible width allowed us to identify the number fluorophores contributing the spectra, their approximate spectra, to perform a statistical analysis of the data and its interpretation.