Online analysis method for intrinsic signal optical imaging

The intrinsic optical imaging technique has been widely applied for the visualization of functional maps in the sensory cortices of mammals. Many current studies refer this mapping in order to focus thereafter on particular features, at some particular locations: a fast and accurate mapping is therefore required. However, even during a successful experiment, the recorded raw data are usually contaminated by some kinds of noise that cannot necessarily be averaged out over the trials. An adequate image data analysis method has to be applied to extract signals closely related neural activities in response to presented stimuli. Thus far two different analysis methods could be adopted: the band-pass filtering and the GIF method [Yokoo T, Knight BW, Sirovich L. An optimization approach to signal extraction from noisy multivariate data. NeuroImage 2001:14;1309–26]. While the latter one is very efficient but requires the whole data in order to maximize the signal to noise ratio, the simple band-pass filtering technically reaches its limits very quickly. Here we propose another filtering method based on the polynomial subtraction of spatially smoothly modulated components. This simple method can visualize well-organized iso-orientation domains of the cat visual cortex with reliability similar to more sophisticated ones while allowing an online visualization of the clean data.