Operation of Network Dynamics in Cultured Hippocampal Neurons on a Multi-electrode Array

Neurons reorganized complex networks on a culture dish and it seems that the living neuronal network can perform certain type of information processing. We showed that dissociated rat hippocampal neurons on a multielectrode array dish formed heterogeneous networks of functional connections. We estimated the functional connections of all combinations of pairs of neurons on the electrodes and represented them into a "connection map". The connection map revealed that each culture contained some hub-like neurons with many functional connections. This cultured neuronal network has unique functional structures which are suitable for perform information processing and the network can re-organize functional connections between neurons in the network. So, it seems that the system can perform some intelligent behavior. We are developing the living brain robot system with premised rules which are corresponding to genetically provided interface of neural network to peripheral system. The premised rules are described by fuzzy logic and they play a role to generate instinctive behavior. The robot can avoid collision by sensor inputs from robot body to neuronal networks, and by output command from the network to motors of the robot. Now we are analyzing the change of connection map moderated by the interaction to outer world via moving robot.