Jennifer D. Simonotto

Uncovering non-linear structure in human ECG recordings

We employ surrogate data techniques and a new correlation dimension estimation algorithm, the Gaussian kernel algorithm, to uncover non-linearity in human electrocardiogram recordings during normal (sinus) rhythm, ventricular tachycardia (VT) and ventricular fibrillation (VF). We conclude that all three rhythms are not linear (i.e. distinct from a monotonic non-linear transformation of linearly filtered noise) and have significant correlations over a period greater than the inter-beat interval. Furthermore, we observe that sinus rhythm and VT exhibit a correlation dimension of approximately 2.3 and 2.4, respectively. The correlation dimension of VF exceeds 3.2. The entropy of sinus rhythm, VT and VF is approximately 0.69, 0.55, and 0.67 nats/s, respectively. These results indicate that techniques from non-linear dynamical systems theory should help us understand the mechanism underlying ventricular arrhythmia, and that these rhythms are likely to be a combination of low dimensional chaos and noise. 2002 Elsevier Science Ltd. All rights reserved.

Automatic identification and recording of cardiac arrhythmia

ECG waveform data showing the spontaneous evolution of ventricular fibrillation (VF) together with its precursors in humans is rare. When such data has been obtained, the resolution is often poor, or the length of pre-onset recording is limited. We describe a new automatic data collection facility that is capable of recording such data. We designed computer software that, in conjunction with Hewlett-Packard proprietary hardware, allows continuous monitoring of physiological waveforms from up to 24 separate hospital beds. Episodes of cardiac arrhythmia (including ventricular tachycardia and ventricular fibrillation) are identified online by either power spectral analysis or nonlinear complexity algorithm. Each episode is automatically recorded for 20 minutes with 10 minutes both before and after onset of the arrhythmia. When monitoring a 6-bed coronary care unit this facility will typically collect around 10-50 recordings per week. The majority of these will be artifacts. However, 2-8 genuine VF episodes will also be recorded per month.

Nonlinear analysis of human ECG during sinus rhythm and arrhythmia

Using a new data collection facility we have recorded spontaneous episodes of VF (and other arrhythmia) in human subjects. We apply correlation dimension analysis and linear and nonlinear surrogate techniques to these data sets. We conclude that human VF is distinct from linearly filtered noise, but not distinct from a noise driven nonlinear system. The correlation dimension estimate from spontaneous VF is between 5 and 8. Furthermore, we show that the temporal complexity (dimensionality) during VF is higher than that observed during VT (4-5) and that during sinus rhythm (3-4). Finally, the characteristic shape of dimension curves obtained during sinus rhythm and arrhythmia are clearly distinct. These results are consistent with those obtained from animal experiments and with computational simulations of spiral wave break-up.

Observation of coherence resonance in stimulated Brillouin scattering

This study provides first experimental evidence of coherence resonance (CR) in an autonomous noise generating process, namely stimulated Brillouin scattering (SBS). A single mode continuous-wave Nd:YAG laser is used as pump to excite SBS in a single mode optical fibre of length 240 meters with very weak feedback. The scattered signal displays stochastic dynamic behaviour from its onset. On increase of the pump signal a window of periodicity appears, emerging from and subsequently subsiding into the stochastic emission. These features are characteristic of CR. This work also shows the dependence of the measure of coherence /spl beta/ (essentially signal-to-noise ratio) as a function of the strength of stochastic emission for a full range of this data. The authors argue through theory that the source of the observed coherence is associated with a transient relaxation oscillation of SBS due to pump depletion in a medium of finite length. For continuous-wave pumping, as in the case of these experiments, the stochastic SBS emission parametrically modulates the pump thereby inducing this relaxation oscillation, which may be sustained for a particular strength of the stochastic Stokes signal.

Coherence Resonance in Stimulated Brillouin Scattering

We examine data from a stimulated Brillouin scattering experiment with very small feedback (<<1%), from a single‐mode optical fiber with a CW YAG laser, and find evidence of Coherence Resonance. For low laser pump levels, the signal is stochastic in nature, as shown by analysis with local dispersion and surrogate data. As the pump level increases, a periodic signal emerges, corresponding to the round trip time of light in the fiber; at higher pump levels this periodic signal is lost, and the data once again becomes stochastic. These findings raise fundamental issues on the emergence of deterministic signals from seemingly stochastic behaviour in nonlinear optical interactions.