Login / Signup

A graphical user interface (NWUSA) for Raman spectral processing, analysis and feature recognition.

Dongliang SongYishen ChenJie LiHaifeng WangTian NingShuang Wang
Published in: Journal of biophotonics (2021)
It is a practical necessity for non-professional users to interpret biologically derived Raman spectral information for obtaining accurate and reliable analytical results. An integrated Raman spectral analysis software (NWUSA) was developed for spectral processing, analysis, and feature recognition. It provides a user-friendly graphical interface to perform the following preprocessing tasks: spectral range selection, cosmic ray removal, polynomial fitting based background subtraction, Savitzky-Golay smoothing, area-under-curve normalization, mean-centered procedure, as well as multivariate analysis algorithms including principal component analysis (PCA), linear discriminant analysis, partial least squares-discriminant analysis, support vector machine (SVM), and PCA-SVM. A spectral dataset obtained from two different samples was utilized to evaluate the performance of the developed software, which demonstrated that the analysis software can quickly and accurately achieve functional requirements in spectral data processing and feature recognition. Besides, the open-source software can not only be customized with more novel functional modules to suit the specific needs, but also benefit many Raman based investigations, especially for clinical usages.
Keyphrases
  • optical coherence tomography
  • machine learning
  • deep learning
  • healthcare
  • magnetic resonance imaging
  • social media
  • data analysis
  • health information
  • liquid chromatography