Blind nonlinear system identification during ventricular fibrillation

A ´blind´ nonlinear system identification procedure for instantaneous nonlinear systems is described. The word ´blind´ is used to emphasize that the procedure is performed when no access to the driving function is available, that is, access is to the output signal only. The problem is important as there are often problems where one cannot apply a test signal yet one wishes to identify the system. In particular an example is presented using data recorded from the myocardium during ventricular fibrillation. The original contributions of the work are the analytical solution for the blind nonlinear system identification procedure presented with underlying assumptions and the application to electrical signals from the heart during ventricular fibrillation.