Feature-based recognition of surface-enhanced Raman spectra for biological targets.

We propose and compare multiple approaches to automatically process data measured through surface-enhanced Raman scattering (SERS), in the context of intracellular molecule probing. It relies on locally detecting the most relevant spectra to retrieve all data independently through indexing, thus avoiding any pre-filtering which occurs with standard processing methods. We first assess our approach on simulated data of the spectrum of Rhodamine 6G, and then validate high-performing methods on experimental measurements of this compound. The optimized method is then utilized to extract and classify the complex SERS response behavior of gold nanoparticles taken in live cells.