Lyapunov exponent measuring device of the finger plethysmogram - Real time chaos monitor

Chaos is a phenomenon in which complex and irregular behavior result from deterministic laws. Recently, it has been suggested that some types of time series from clinical monitors which were previously considered to be irregular is actually chaotic. We developed the device for real time chaotic analysis of the finger plethysmogram. The finger plethysmogram was obtained from the sensor for the infrared emission transmission pulse oximeter. The analog time series of finger plethysmogram were converted into the digital data with a 12 bit analog/digital converter. This digital data was transferred to a personal computer via a USB (universal serial bus) connection. The chaotic analysis was done with the personal computer. The attractor was restructured with time series by Takens' delay embedding theorem (Fig. 1). In this device, the data sampling rate was 5 msec, τ was 50 msec, and the embedding number of dimensions were four-dimensional. The Lyapunov exponent that was the indication of the variability in the attractor orbit was calculated using Sano's method. In this device, the attractor was restructured with the time series of the finger plethysmogram of 17.5 sec, and the maximum Lyapunov exponent of each 1 sec was calculated. Fig. 2 shows the change in the Lyapunov exponent measured by this device during the removal of internal fixation implant plates in a 20-year-old female under propofol-nitrous oxide anesthesia. It is shown that the Lyapunov exponent changed in each situation: (A) from the monitor installation to the vein securing, (B) from the anesthesia induction to the start of operation, (C) from the start to the end of operation, and (D) from the end of operation to the extubation and emergence. In future work, we plan to observe the change of the Lyapunov exponent during general anesthesia and intravenous sedation using this device, and to clarify the change in the characterization of chaos in these situations.