Unmixing Hyperspectral SRS Images in the Cell-Silent Region of the Raman Spectrum Using Phasor Analysis.

Hyperspectral stimulated Raman scattering (SRS) microscopy is rapidly becoming an established method for chemical and biomedical imaging due to the combination of high spatial resolution and chemical information contained within the three-dimensional data set. Chemometric analysis techniques based on linear unmixing, or multivariate analysis, have become indispensable when visualizing hyperspectral data sets. The application of spectral phasor analysis has also been extremely fruitful in this regard, providing a convenient method to retrieve the spatial and chemical components of the data set. Here, we demonstrate the application of spectral phasor analysis for unmixing the overlapping spectral features within the cell-silent region of the SRS spectrum (2000-2300 cm^-1). In doing so, we show it is possible to identify specific Raman signals for DNA, proteins, and lipids following glucose-d₇ metabolism in dividing cells. In addition, we show that spectral phasor analysis is capable of distinguishing different bioorthogonal Raman signals including alkynes and carbon-deuterium (C-D) bonds. We demonstrate the application of spectral phasor analysis for multicomponent unmixing of bioorthogonal Raman groups for high-content cellular imaging applications.