Fast and robust identification of single bacteria in environmental matrices by Raman spectroscopy

Abstract

We report on our recent results on robust identification of single bacterial cells embedded in various environments using Spontaneous Raman Scattering. Five species of bacteria were considered, two of which (B. Subtilis and E. Coli) were grown under various conditions, or embedded in two real-world matrices. We recorded the Raman spectra of single cells with a confocal instrument developed in our lab, and performed identification at the species level. Our system integrates a Lensfree imaging module that allows fast detection of bacteria over a large Field-Of-View. Identification rates comparable to those obtained on lab cultures were possible using a comprehensive database containing spectra from bacteria in all environments. In addition, B. Subtilis was correctly identified in 95.5% of the cases using a database composed exclusively of spectra obtained in standard conditions. This is very promising for pathogen threat detection where the construction of an exhaustive database may be challenging.