Electronic sensor array incorporating artificial neural network algorithms for rapid identification and quantification of Escherichia coli and Salmonella enterica serovar Typhimurium and their volatile metabolites

A rapid method to identify and quantify E. coli and Salmonella typhimurium and their specific volatile metabolites in nutrient media as well as in packaged alfalfa sprouts was developed using an array of 12 nonspecific metal oxide electronic sensors incorporating artificial neural network algorithms. The metabolic volatile compounds used as indicators of E. coli and Salmonella typhimurium in the samples were identified using solid phase microextraction coupled with gas chromatograph/mass spectrometer (SPME/GC/MS). Principal Component Analysis (PCA) was used for data exploration and dimensional reduction. The Mahalanobis distance metric was determined based on Discriminant Factor Analysis (DFA) for sample classification to differentiate volatiles in control samples from that containing the target microorganisms. Artificial neural networks were trained to identify and quantify E. coli and Salmonella typhimurium and their volatile metabolites.; In all studies, the neural networks were shown to be capable of correlating voltametric responses with number of E. coli, the relationship between responses and concentrations of volatile metabolites, and the concentrations of individual components from mixtures with low mean square errors. The electronic sensor array was found to be satisfactorily correlated with colony counting and GC/MS methods. This technique provides a rapid, simple, and precise analysis of the biochemical composition of microbiological systems, for identification of potentially pathogenic microorganisms.