E-nose and E-tongue systems for the Bacteria Detection in milk

Abstract

The purpose of this paper is to describe the combination of an E-nose and E-tongue that were evaluated for the E. coli detection at different concentrations, as well as their ability to discriminate this bacterium from others, such as Klebsiella pneumoniae and Salmonella enterica in pasteurized milk. In this study, gold and carbon electrodes were tested in the E-tongue. For data processing, multivariate analysis techniques were used to discriminate the measurements, where the Principal Components Analysis (PCA) and Linear Discriminant Analysis (LDA) methods were applied. Likewise, for the data classification, the Vector Support Machines (SVM) through the linear kernel and Radial Basis Function (RBF) algorithms were used, and the same way as the k-Nearest Neighbor (k-NN) method. When evaluating the capacity of the proposed methodologies to detect and classify E. coli, S. enterica, and K. pneumoniae in pasteurized milk, it was observed that both the E-nose (TGS 826 sensor) and the E-tongue (gold electrode) obtained comparable results with 94.7% and 92.5% success rate respectively. Both devices successfully detected and classified the three bacteria tested, clearly differentiating them from the sterile milk samples. On the other hand, the electronic tongue with a gold electrode achieved a 98.7% success rate in the discrimination of decreasing concentrations of E. coli, from 1x 06 CFU/ml to 1x10-2 CFU/ml, in pasteurized milk.