Abstract
Methods of human bacterial pathogen identification need to be fast, reliable, inexpensive, and time efficient. These requirements may be met by vibrational spectroscopic techniques. The method that is most often used for bacterial detection and identification is Fourier transform infrared spectroscopy (FTIR). It enables biochemical scans of whole bacterial cells or parts thereof at infrared frequencies (4,000-600 cm-1). The recorded spectra must be subsequently transformed in order to minimize data variability and to amplify the chemically-based spectral differences in order to facilitate spectra interpretation and analysis. In the next step, the transformed spectra are analyzed by data reduction tools, regression techniques, and classification methods. Chemometric analysis of FTIR spectra is a basic technique for discriminating between bacteria at the genus, species, and clonal levels. Examples of bacterial pathogen identification and methods of differentiation up to the clonal level, based on infrared spectroscopy, are presented below.
Keywords: Artificial neural network (ANN), bacterial pathogen detection, chemometrics, fourier transform infrared spectroscopy (FTIR).