Extracting 1/f fluctuation from the arterial blood pressure of an artificial heart

We have studied the fluctuations of an artificial circulation for the analysis of the physiological aspects; however, the conventionally used fast Fourier transform (FFT) method cannot separate harmonic oscillations, such as respiratory and Mayer waves, from the 1/f fluctuation, which has been though to represent underlying fractal dynamics. Fractal structure was shown in the strange attractor with chaotic dynamics, which is thought to be a flexible and intelligent system. In this study, the coarse-graining spectral analyzing (CGSA) method was utilized to quantitatively evaluate the proportion of the 1/f fluctuation in the total power in the frequency domain and to analyze artificial circulation in the whole system. We implanted two pneumatically actuated ventricular assist devices as biventricular bypasses (BVBs) in chronic animal experiments using 4 healthy adult goats. To compare the natural and prosthetic circulation of each experimental animal, the BVB-type complete prosthetic circulation model with electrically induced ventricular fibrillation was adopted. All hemodynamic parameters of natural and prosthetic circulation were recorded under awake conditions and calculated with the use of a personal computer. With the use of the CGSA method, time-series data of the hemodynamics were analyzed and fractal percentages, extracting the 1/f fluctuation from a given time series, were calculated. Fractal percentages of the arterial blood pressure were 85.8 +/- 10.7% and 82.0 +/- 7.3% with natural and artificial circulation, respectively (not significant [NS]). 1/f fluctuation showed the characteristics of being fractal in a time series. The fractal structure showed robustness and error resistance in nonlinear dynamics. Therefore, our results suggest that the circulatory regulatory system of the artificial heart may have desirable characteristics such as error resistance.