Characterizing nonlinearity in ventricular fibrillation

The authors apply techniques from nonlinear dynamical systems theory to determine whether Ventricular Fibrillation (VF) comes from a stochastic, or a more complex nonlinear process. Using electrocardiographic (ECG) data from seven pigs during ventricular fibrillation the authors apply various surrogate data techniques to test the observed data against a series of hypotheses. Standard linear surrogate techniques demonstrate that VF is inconsistent with data generated by a monotonic nonlinear transformation of a linearly filtered noise source. Nonlinear modeling methods show that VF is consistent with the output of a system exhibiting a stable periodic orbit and small scale high dimensional structure (either noise or chaos)