Dealing with in vivo multi-electrode array data

Our main aim of the research is to develop the first user-friendly software package to deal with in vivo multi-electrode array (MEA) data, as an essential tool and stage towards understanding how an odour is coded by the olfactory bulb(OB), how an image is coded by the visual systems, and finally how multi-regions in the nervous systems work coherently. We expect our adventures will provide a complete answer to some fundamental questions related to the information processing in the OB, and shed new lights onto the related issues in the brain in general. Use of the MEA techniques to detect signals coded by populations of neurons will be important not only for elucidating mechanisms of brain function, and for alleviating medical problems such as paralysis. Since this sort of work may also guide development of intelligent machines it is being used to provide data for computer simulations of brain activity.

The framework for developing statistical software, applying them to tackling MEA recordings and developing models to account for the information processing in the olfactory bulb and in the brain in general is important for opening up applications to biology and engineering. There is mounting experimental evidence to support the importance of information processing in spatio-temporal patterns. There is also increasing use of the MEA techniques in the development of brain-machine interfaces. Engineering and AI research can also benefit a lot from recent results revealed in fundamental neuroscience experiments, in particular a combination of experiment and theory. It seems we are one of the first groups aiming to develop and actually apply statistics for tackling the MEA recordings, in combination with experiments and modelling. This timely and novel approach can provide a generic framework for applications both in biology and in engineering. Our approach is applicable to other areas where complex experimental data are available such as Microarray gene data and fMRI.