Transmission of spatio-temporal patterns from biological to artificial neural networks by a multi-electrode array

The monitoring of a set of individual neurons in cultured biological networks or in the brain has become feasible with the used/development of multi-electrode arrays (MEA). However, even with the huge mass of data, getting suitable information about the actual spatio-temporal context of the analyzed biological network is not easy. In this paper we present a new conception and first results of analyzing the measured data by a recurrent artificial neural network with similar parameters as the biological network. The signals of the biological network transfer into the artificial one and the balanced artificial network becomes a part of the dynamics of the biological network. The artificial network is more transparent for advanced methods to analyze synchronous firing patterns (i.e., polychronization) and may also generate adequate feedback signals to the biological network for using as a recurrent neurointerface.