Theoretical signals (sine wave, Henon attractor, and white noise) used for the parameter optimization of the PPE estimation. Discrete time series (a) and their two-dimensional representation in the phase space phase plot (b) are shown.
Influence of the evolving time (a) and the embedding dimension (b) on the estimation of the LLE for the Henon attractor. The squared errors of the difference between the real and estimated LLEs are shown in panel (c) (combined examination of both parameters).
Investigation of the transitions between sine wave and Henon attractor (a) as well as sine wave and white noise (b), respectively. Displayed are the PPEs of the unaffected signals (light gray black lines) and the PPE of the transition between signals (dark gray line).
Influence of crucial parameters on the estimation of the PPE for fetal ECoG during NREM and REM sleep at 130 dGA.
Results of the test of nonlinearities in the fetal ECoG at 106 and 130 dGA. The mPPE of the original ECoG and the respective linear AR modeling during NREM and REM sleep at 130 dGA or cortical activation and deactivation periods at 106 dGA of all investigated fetuses are shown (, mean deviation).
Time course of the original recorded ECoG and the estimated SEF, nonlinear stability PPE, and parametric mBA at 106 and 130 dGA. An example of a 60 min ECoG epoch of one fetus is shown. Gray boxes indicate distinguished periods of cortical deactivation at 106 dGA or NREM sleep at 130 dGA, respectively.
Comparison of the mPPE (a) and mSEF (b) during the different identified sleep states at 106 and 130 dGA of all investigated fetuses (, mean deviation).
Article metrics loading...
Full text loading...