Correlation integral analysis of epicardial surface potentials derived from isolated rat hearts

Driving an isolated rat heart with external, periodic electrical stimuli results in electrical responses of the heart, recorded as epicardial surface potentials (ESP). The ESP signal is a recurrent signal, yet not perfectly periodic. The electrical system of the heart underlying such variability is investigated. ESP complexes are fitted by orthogonal Chebyshev polynomials resulting in a series of coefficients (CC) representing the data. Randomized surrogate data (SD) for ESP and CC are created. Correlation integrals, C(r), are calculated for ESP data, CC, representing the data, and their SD. The C(r) of the four data sets are compared. Results show that C(r) of ESP and CC are similar, while both differ from the C(r) of their SD. These results imply that the system underlying the ESP signal differs from an infinite dimensional system, and may be a finite dimensional dynamical system of a dimension around 5.