Application of artificial neural network for the identification of fresh water bacteria

A method based on artificial neural network (ANN) for monitoring aquatic bacteria which would be useful for health care is presented. Environmental micro-organisms include a large number of taxa. Some species that normally are not pathogenic can represent a risk in certain conditions, such as old people and immuno-compromised individuals. A system based on unsupervised ANN has been set up using the fatty acid profiles of standard strains, obtained by gas-chromatography, as learning data. The Kohonen output map resulted in a powerful tool for identification of fresh isolates coming from a line of the major civil water system of Genova (Italy).