In the present research the dynamics of cardiac activity is investigated by means of exploring topological properties of the signal attractor in a phase space with dimension m and stability of a spiral wave structure commonly observed in the medium of heart muscle and obtained from reaction-diffusion equations at its computational modeling.Evolution of cardiac activity is investigated by means of nonlinear dynamics, namely the method of temporal localization on the attractor reconstructed from a digitized ECG signal.Convergence for the function of topological instability at changing dimensionality is proven both theoretically and numerically, independently from personal features of subjects in a latter case.This provides an opportunity to estimate the complexity (expressed through the number of freedom degrees) of cardiac dynamics.On the other hand, this instability function normalized by its average displays a different kind of behavior that somewhat differs for various persons and reflects their individual features.The essential reduction of computation time and necessary statistics are also attained by means of the developed algorithm.These results provide information that reflects temporal evolution of cardiac activity and therefore can be used for the purposes of early diagnosis.