Figure 2

Artificial neural network (ANN) for the identification of Streptococcus spp. and related cocci. Hierarchical structure for the modular artificial neural network (ANN) for the identification of Streptococcus spp. and related cocci. On the first level of the ANN architecture, streptococci are divided from other gram-positive, catalase-negative cocci. On the second level, A. viridans, Enterococcus spp., Lactococcus spp., S. agalactiae, S. canis, outgroup I could be identified. The other three levels were used for identifying S. bovis, S. dysgalactiae, S. parauberis, S. pyogenes, S. uberis and outgroup II. Species contained in outgroup I and II were used as negative controls and are not associated to bovine mastitis. The following strains were used for outgroup I: S. mitis, S. pneumoniae, S. sanguinis, S. porcinus, S. gallinarum and for outgroup II: Streptococcus equi, Streptococcus gallolyticus subsp. macedonicus, Streptococcus suis.