Improvement of surface-enhanced Raman scattering detection and imaging by multivariate curve resolution methods

Surface-enhanced Raman scattering (SERS) tags have been intensively applied in biological detection and imaging in recent years. However, both applications often suffer from high Raman background signals from containers such as 96-well plates or autofluorescence signals from biological tissues. Here, we greatly mitigate the influence of these high background Raman or fluorescent signals in both biological detection and imaging using two multivariate curve resolution (MCR) methods including negative matrix factorization and classical least squares. The limit of detection is lowered by one order of magnitude after applying MCR methods to detect target SERS tags in a 96-well plate. Additionally, in a multiplexed cell imaging assay, both false-negative and false-positive results were eliminated with the aid of MCR methods. Accordingly, we suggest a wider application of MCR methods during both biological detection and imaging of SERS tags with high background signals.