The possibility of kohonen self-organizing map applications in medicine

In many instances, artificial neural networks (ANN) are used as interfaces between real situations with real data and machines or computers. Often, situations arise in which the real environment cannot be eliminated, among them data collection and medical data processing. This contribution deals with several medical domains in which ANNs prove to be a very good alternative to standard methods. The text of the contribution starts from a research project in which we cooperate with several medical facilities. At the centre of concern in our research orientation is improvement of diagnostics, and the ensuing improvement in the treatment of both children and adult patients suffering from diseases affecting social interaction. In all cases described, we apply one of the ANN methods, with particular attention given to alternatives to self-organizing maps (SOM).