A Nonlinear Filtering Technique for the Identification of Biological Signals

A new nonlinear filtering technique is presented allowing extraction of relevant biological signals from background. The technique s derived from the one of dc restoration used in nuclear electronics. A discussion is given of the problem and a solution through the use of a nonlinear filter is proposed; the filter's way of operating is then described together with a complete multiunit neural spike processing instrument using this filtering technique. Some experimental results are presented.