Signal to noise ratio of Raman spectra of biological samples

摘要

Raman microspectroscopy is an optoelectronic technique that can be used to evaluate the chemical composition of biological samples. Raman spectroscopy has been shown to be a powerful classification tool for the investigation of various cancer related diseases including bladder, breast and cervical cancer. Raman scattering is an inherently weak process with approximately 1 in 107 photons undergoing scattering. For this reason, noise from the recording system can have a significant impact on the quality of the signal and its suitability for classification. Different camera settings when obtaining spectra from charge-coupled devices can result in significantly different noise performance. This paper provides an investigation into practical aspects of retrieving the signal from the charge-coupled device, and the effects of integration time and multiple acquisitions on the signal to noise ratio of Raman spectra, with a particular focus on biological samples. The main sources of noise are shot noise, CCD dark current, readout noise, and cosmic ray artefacts. Shot noise and dark current noise are time dependent and so there are practical considerations when choosing an integration time. Readout noise is inherent in each individual recording, which may be compounded when averaging spectra together. Our results demonstrate that read parameters and read modes can greatly influence the signal to noise ratio. We also discuss experimental conditions and processing methods that can mitigate these effects.