Combination of laser-induced breakdown spectroscopy and Raman spectroscopy for multivariate classification of bacteria ☆ D. Prochazka a,b, ⁎, M. Mazura a , O. Samek c , K. Rebrošová d , P. Pořízka a,b , J. Klus a,b , P. Prochazková e , J. Novotný a,b , K. Novotný a,b , J. Kaiser a,b a Central European Institute of Technology, Brno University of Technology, Purkyňova 123, CZ 61200, Brno, Czech Republic b AtomTrace a.s., Kolejní 9, 61200 Brno, Czech Republic c I Institute of Scientific Instruments of the CAS, Královopolská 147, Brno 61264, Czech Republic d Department of Microbiology, Faculty of Medicine and St. Anne's Faculty Hospital, Brno 65691, Czech Republic e Masaryk University, Faculty of Science, Department of Chemistry, Kamenice 735/5, 625 00 Brno, Czech Republic abstract article info Article history: Received 28 February 2017 Received in revised form 2 November 2017 Accepted 9 November 2017 Available online 10 November 2017 In this work, we investigate the impact of data provided by complementary laser-based spectroscopic methods on multivariate classification accuracy. Discrimination and classification of five Staphylococcus bacterial strains and one strain of Escherichia coli is presented. The technique that we used for measurements is a combination of Raman spectroscopy and Laser-Induced Breakdown Spectroscopy (LIBS). Obtained spectroscopic data were then processed using Multivariate Data Analysis algorithms. Principal Components Analysis (PCA) was selected as the most suitable technique for visualization of bacterial strains data. To classify the bacterial strains, we used Neural Networks, namely a supervised version of Kohonen's self-organizing maps (SOM). We were process- ing results in three different ways - separately from LIBS measurements, from Raman measurements, and we also merged data from both mentioned methods. The three types of results were then compared. By applying the PCA to Raman spectroscopy data, we observed that two bacterial strains were fully distinguished from the rest of the data set. In the case of LIBS data, three bacterial strains were fully discriminated. Using a com- bination of data from both methods, we achieved the complete discrimination of all bacterial strains. All the data were classified with a high success rate using SOM algorithm. The most accurate classification was obtained using a combination of data from both techniques. The classification accuracy varied, depending on specific samples and techniques. As for LIBS, the classification accuracy ranged from 45% to 100%, as for Raman Spectroscopy from 50% to 100% and in case of merged data, all samples were classified correctly. Based on the results of the experiments presented in this work, we can assume that the combination of Raman spectroscopy and LIBS significantly enhances discrimination and classification accuracy of bacterial species and strains. The reason is the complementarity in obtained chemical information while using these two methods. © 2017 Elsevier B.V. All rights reserved. Keywords: Laser-induced breakdown spectroscopy Raman spectroscopy Chemometrics Bacteria 1. Introduction A conventional identification of bacteria requires some knowledge of their morphological, biochemical, physiological and genetic charac- teristics. A bacteria identification procedure consists of bacteria cultiva- tion, separation, its study under an optical microscope, etc. The whole procedure is time-consuming, inefficient and requires an expertise of a trained microbiologist. Therefore there is an effort to develop a faster and automatic method for bacteria identification. One of the promising methods for biological samples identification is Laser-Induced Breakdown Spectroscopy (LIBS) [1–6]. The potential of LIBS for bacteria study and identification is well demonstrated in several publications. Here will be mentioned only several manuscripts, interest- ing from the point of view of goals of this work. S. Morel et al. [7] per- formed discrimination of bacteria species and strains based on the cumulative spectral line intensities ratio. It was presented that this ap- proach works well, however, in the case of vast datasets it is still time- consuming and inefficient. M. Baudelet et al. [8] studied the influence of laser pulse duration on analytical performance of LIBS for bacteria identification. In the manuscript it is shown that in a direct comparison of nanosecond and femtosecond laser pulse the femtosecond has less interference with emissions from the ambient air. Improved figures of merit can be achieved by employing multivari- ate data analysis (MVDA, in spectroscopy often called chemometrics). Spectrochimica Acta Part B 139 (2018) 6–12 ☆ Selected paper from the 9th International Conference on Laser-Induced Breakdown Spectroscopy (LIBS), Chamonix-Mont-Blanc, France, September 12 – September 16 2016. ⁎ Corresponding author at: Central European Institute of Technology, Brno University of Technology, Purkyňova 123, CZ-61200 Brno, Czech Republic. E-mail address: david.prochazka@ceitec.vutbr.cz (D. Prochazka). https://doi.org/10.1016/j.sab.2017.11.004 0584-8547/© 2017 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Spectrochimica Acta Part B journal homepage: www.elsevier.com/locate/sab