Microelectrodes for monitoring extracellular neuronal activity
Microelectrode arrays (MEAs) have become a valuable research tool for studying the in-vitro electrophysiological activity of excitable and spontaneously active cells, such as neurons, in simple networks. An arrangement of several electrodes allows targeting several sites for non-invasive simulation and extracellular recording of the electrophysiological activity at once. Despite their simplified level of organisation, these ex vivo neurobiological systems can form a bi-dimensional physical model of the brain which preserves functional behaviours of its three dimensional structure. The emerging electrophysiological patterns range from apparent stochastic spiking to organised bursting and the structure of individual neurons is strongly influenced by the interaction with the surroundings. NEUROBIT project aimed to develop algorithms and techniques for establishing a bi-directional connection between in vitro cultured populations of neurons and external devices. The main objective was to enable the embodiment of cultured neurons in an actual physical body and the instruction of the hybrid system's biological component to process information in a goal-oriented way. Rather than using commercially available tools with limited compatibility with the different experimental recording techniques, project partners at Telecom Italia Learning Services S.p.A. endeavoured to develop a more suitable software package. It integrates features that permit easy customisation for different types of microelectrode arrays and A/D boards, including data transmission through several protocols and devices. Induced changes in the neuronal network activity are identified as parameters changes among different experimental phases, while the experimental setup features are managed through a friendly Graphical User Interface (GUI). In order to reduce the amount of raw data to be processed, it was desirable to develop peak detection algorithms that would extract distinct collective functional states for further analysis. The software architecture permits the user to easily add new functionalities and further processing algorithms through modular structures, called Dynamic Link Libraries (DLL). This is considered as an essential step towards the realisation of complex software tools for the real-time and automatic processing of electrophysiological data from large populations of neurons.
