Modeling spiking activity of in vitro neuronal networks through non linear methods

Neuroscience research is even more exploiting technologies developed for electronic engineering use: this is the case of Micro-Electrode Array (MEA) technology, an instrumentation which is able to acquire in vitro neuron spiking activity from a finite number of channels. In this work we present three models of synaptic neuronal network connections, called “Full-Connected”, “Hierarchical” and “Closed-Path”. Related to each one we implemented an index giving quantitative measures of similarity and of statistical dependence among neuron activities recorded in different MEA channels. They are based on Information Theory techniques as Mutual and Multi Information: the last one extending the pair-wise information to higher-order connections on the entire MEA neuronal network. We calculated indexes for each model in order to test the presence of self-synchronization among neurons evolving in time, in response to external stimuli such as the application of chemical neuron-inhibitors. The availability of such different models helps us to investigate also how much the synaptic connections are spatially sparse or hierarchically structured and finally how much of the information exchanged on the neuronal network is regulated by higher-order correlations.