Spectral pre-processing for biomedical vibrational spectroscopy and microspectroscopic imaging

Abstract

Recent years have seen substantial progress toward the application of infrared (IR) and Raman spectroscopy as useful analytical tools in biomedical research. Vibrational spectroscopy, and (in particular) microspectroscopy, have been successfully applied to biomedical samples ranging from intact microorganisms, eukaryotic cells, body fluids, and tissues. The progress in the field was driven not only by technical developments but also by the effective implementation of modern concepts of spectral analysis. Pre-processing has been identified as an indispensable part of spectral data analysis. It involves, among others, outlier rejection, normalization, filtering, detrending, transformation, folding and feature selection. Goals of spectral pre-processing include better interpretability of the spectra, higher robustness and improved accuracy of subsequent quantitative or classification analysis. The aim of this review article is to explore the concepts and techniques of a variety of individual pre-processing methods and to discuss the applicability of different pre-processing techniques in the context of practical applications of biomedical IR or Raman spectroscopy. It is hoped that this article not only represents a useful guideline for beginners in the field of biomedical applications of vibrational spectroscopy, but serves also as a source of reference for more experienced spectroscopists.