Multivariate approaches for SERS data analysis in clinical applications

Surface-enhanced Raman spectroscopy (SERS) has attracted great attention in the field of nanomedicine due to its ultrasensitive detection capability. When combined with efficient multivariate statistical methods, SERS is emerging as a novel and powerful tool for biosample analysis toward applications from species identification to clinical diagnosis. Spectral data analysis is essential for exploring potential diagnostic information and plays an indispensable role in SERS-based assay. In this chapter, we will introduce the basic concepts and functional descriptions of multivariate data analysis approaches, including unsupervised and supervised statistical algorithms. Besides, recent advances on practical applications of these statistical analysis strategies for label-free and labeling SERS assay in biological and clinical detection will be summarized and discussed.