Blind Separation of Neuronal Activities from Optical Imaging Data

Abstract

The optical imaging method is one of the most useful techniques to investigate dynamic neuronal activities of the brain. Available forms of the method, however, require averaging of the optical records tens of times. A new version presented here has been developed by application of the Independent Component Analysis (ICA), and it is devised to circumvent the time-consuming averaging process. As a preliminary work, parameters associated with ICA were determined on the basis of a computer simulation involving membrane potentials and noise signals. The method was then applied to the optical data obtained from hippocampal slices. The neuronal activities extracted by the present technique are in good agreement with the results produced by conventional technique that needs extensive averaging.