An integrated software system for processing, analyzing, and classifying Raman spectra

Abstract

Raman spectroscopy is a powerful technique for analyzing various substances, but a significant amount of processing is required to extract useful information from raw spectra. Many research groups develop custom software systems to process spectra, which leads to standard algorithms being repeatedly implemented and novel techniques being difficult to share. This paper presents an integrated software system that processes, analyzes, and classifies Raman spectra. The system is freely available ( http://cares.wayne.edu/rp/), open, cross-platform, and extensible, enabling interested parties to utilize and contribute to a common implementation of Raman processing algorithms. A wide variety of useful features are included, such as noise filtering, automated background fluorescence subtraction, principal component analysis, linear discriminant analysis, artificial neural network classification, and support vector machine classification. The system supports customizable groupings of related spectra and is able to retrieve spectra using a searchable database. Everything is executed through a simple graphical user interface. The developed system was used to process two data sets containing different types of spectra. The results demonstrate that the system is capable of fast, accurate processing of Raman spectra. The open system will become even more powerful if other developers are willing to add new and improved features. In addition, it is intended that the freely available system will make Raman spectroscopy more accessible to all researchers and promote progress in the field.