In the last decades, there is a vivid interest of many researchers about algorithms with natural procedure similarities. Artificial Neural Networks (ANNs) are wide known algorithms, which have been developed in order to solve complex combinational problems and can be proven as a powerful tool in medical research. This paper implements computational algorithms, which were based on artificial neural networks for the dementia's types distinguish. Diversity of ANN architectures are implemented and assessed for correct prediction of types of dementia, based on available cases. The obtained results prove that ANNs have the ability to be incorporated in the field of Neurology.